A Critical Analysis of Patient Safety Practices - Medical and Public ...

Evidence Report/Technology Assessment
Number 43
Making Health Care Safer: A Critical
Analysis of Patient Safety Practices
Prepared for:
Agency for Healthcare Research and Quality
U.S. Department of Health and Human Services
2101 East Jefferson Street
Rockville, MD 20852
www.ahrq.gov
Contract No. 290-97-0013
Prepared by:
University of California at San Francisco (UCSF)–Stanford University
Evidence-based Practice Center
Editorial Board
Kaveh G. Shojania, M.D. (UCSF)
Bradford W. Duncan, M.D. (Stanford)
Kathryn M. McDonald, M.M. (Stanford)
Robert M. Wachter, M.D. (UCSF)
Managing Editor
Amy J. Markowitz, JD
Robert M. Wachter, MD
Project Director
Kathryn M. McDonald, MM
UCSF–Stanford EPC Coordinator
AHRQ Publication 01-E058
July 20, 2001 (Revised printing, indexed)

Preface
The Agency for Healthcare Research and Quality (AHRQ), formerly the Agency for
Health Care Policy and Research (AHCPR), through its Evidence-based Practice Centers
(EPCs), sponsors the development of evidence reports and technology assessments to assist
public- and private-sector organizations in their efforts to improve the quality of health care
in the United States. The reports and assessments provide organizations with comprehensive,
science-based information on common, costly medical conditions and new health care
technologies. The EPCs systematically review the relevant scientific literature on topics
assigned to them by AHRQ and conduct additional analyses when appropriate prior to
developing their reports and assessments.
To bring the broadest range of experts into the development of evidence reports and
health technology assessments, AHRQ encourages the EPCs to form partnerships and enter
into collaborations with other medical and research organizations. The EPCs work with these
partner organizations to ensure that the evidence reports and technology assessments they
produce will become building blocks for health care quality improvement projects throughout
the Nation. The reports undergo peer review prior to their release.
AHRQ expects that the EPC evidence reports and technology assessments will inform
individual health plans, providers, and purchasers as well as the health care system as a whole
by providing important information to help improve health care quality.
We welcome written comments on this evidence report. They may be sent to: Director,
Center for Practice and Technology Assessment, Agency for Healthcare Research and Quality,
6010 Executive Blvd., Suite 300, Rockville, MD 20852.
John M. Eisenberg, M.D. Douglas B. Kamerow, M.D.
Director Director, Center for Practice and
Agency for Healthcare Technology Assessment
Research and Quality Agency for Healthcare
Research and Quality
The authors of this report are responsible for its content. Statements in the report should not be
construed as endorsement by the Agency for Healthcare Research and Quality or the U.S.
Department of Health and Human Services of a particular drug, device, test, treatment, or other
clinical service.

Structured Abstract
Objectives: Patient safety has received increased attention in recent years, but mostly with a
focus on the epidemiology of errors and adverse events, rather than on practices that reduce such
events. This project aimed to collect and critically review the existing evidence on practices
relevant to improving patient safety.
Search Strategy and Selection Criteria: Patient safety practices were defined as those that
reduce the risk of adverse events related to exposure to medical care across a range of diagnoses
or conditions. Potential patient safety practices were identified based on preliminary surveys of
the literature and expert consultation. This process resulted in the identification of 79 practices
for review. The practices focused primarily on hospitalized patients, but some involved nursing
home or ambulatory patients. Protocols specified the inclusion criteria for studies and the
structure for evaluation of the evidence regarding each practice. Pertinent studies were identified
using various bibliographic databases (e.g., MEDLINE, PsycINFO, ABI/INFORM, INSPEC),
targeted searches of the Internet, and communication with relevant experts.
Data Collection and Analysis: Included literature consisted of controlled observational
studies, clinical trials and systematic reviews found in the peer-reviewed medical literature,
relevant non-health care literature and “gray literature.” For most practices, the project team
required that the primary outcome consist of a clinical endpoint (i.e., some measure of morbidity
or mortality) or a surrogate outcome with a clear connection to patient morbidity or mortality.
This criterion was relaxed for some practices drawn from the non-health care literature. The
evidence supporting each practice was summarized using a prospectively determined format.
The project team then used a predefined consensus technique to rank the practices according to
the strength of evidence presented in practice summaries. A separate ranking was developed for
research priorities.
Main Results: Practices with the strongest supporting evidence are generally clinical
interventions that decrease the risks associated with hospitalization, critical care, or surgery.
Many patient safety practices drawn primarily from nonmedical fields (e.g., use of simulators,
bar coding, computerized physician order entry, crew resource management) deserve additional
research to elucidate their value in the health care environment. The following 11 practices were
rated most highly in terms of strength of the evidence supporting more widespread
implementation.
• Appropriate use of prophylaxis to prevent venous thromboembolism in patients at risk;
• Use of perioperative beta-blockers in appropriate patients to prevent perioperative
morbidity and mortality;
• Use of maximum sterile barriers while placing central intravenous catheters to prevent
infections;
v

• Appropriate use of antibiotic prophylaxis in surgical patients to prevent postoperative
infections;
• Asking that patients recall and restate what they have been told during the informed
consent process;
• Continuous aspiration of subglottic secretions (CASS) to prevent ventilator-associated
pneumonia;
• Use of pressure relieving bedding materials to prevent pressure ulcers;
• Use of real-time ultrasound guidance during central line insertion to prevent
complications;
• Patient self-management for warfarin (Coumadin) to achieve appropriate outpatient
anticoagulation and prevent complications;
• Appropriate provision of nutrition, with a particular emphasis on early enteral nutrition in
critically ill and surgical patients; and
• Use of antibiotic-impregnated central venous catheters to prevent catheter-related
infections.
Conclusions: An evidence-based approach can help identify practices that are likely to
improve patient safety. Such practices target a diverse array of safety problems. Further research
is needed to fill the substantial gaps in the evidentiary base, particularly with regard to the
generalizability of patient safety practices heretofore tested only in limited settings and to
promising practices drawn from industries outside of health care.
This document is in the public domain and may be used and reprinted without permission
except those copyrighted materials noted for which further reproduction is prohibited without the
specific permission of copyright holders.
Suggested Citation:
Shojania KG, Duncan BW, McDonald KM, et al., eds. Making Health Care Safer: A
Critical Analysis of Patient Safety Practices. Evidence Report/Technology Assessment No. 43
(Prepared by the University of California at San Francisco–Stanford Evidence-based Practice
Center under Contract No. 290-97-0013), AHRQ Publication No. 01-E058, Rockville, MD:
Agency for Healthcare Research and Quality. July 2001.
vi

CONTENTS
Summary.........................................................................................................................1
Evidence Report .............................................................................................................9
PART I. OVERVIEW ..........................................................................................................11
Chapter 1. An Introduction to the Report..............................................................................13
1.1. General Overview .......................................................................................................13
1.2. How to Use this Report ...............................................................................................18
1.3. Acknowledgments.......................................................................................................20
Chapter 2. Drawing on Safety Practices from Outside Health Care......................................25
Chapter 3. Evidence-based Review Methodology ................................................................29
PART II. REPORTING AND RESPONDING TO PATIENT SAFETY PROBLEMS ...........................39
Chapter 4. Incident Reporting ...............................................................................................41
Chapter 5. Root Cause Analysis............................................................................................51
PART III. PATIENT SAFETY PRACTICES & TARGETS ...........................................................57
Section A. Adverse Drug Events (ADEs).................................................................................58
Chapter 6. Computerized Physician Order Entry (CPOE) with Clinical Decision Support
Systems (CDSSs)....................................................................................................................59
Chapter 7. The Clinical Pharmacist’s Role in Preventing Adverse Drug Events..................71
Chapter 8. Computer Adverse Drug Event (ADE) Detection and Alerts..............................79
Chapter 9. Protocols for High-Risk Drugs: Reducing Adverse Drug Events Related to
Anticoagulants........................................................................................................................87
Chapter 10. Unit-Dose Drug Distribution Systems .............................................................101
Chapter 11. Automated Medication Dispensing Devices....................................................111
Section B. Infection Control ...................................................................................................118
Chapter 12. Practices to Improve Handwashing Compliance .............................................119
Chapter 13. Impact of Barrier Precautions in Reducing the Transmission of Serious
Nosocomial Infections..........................................................................................................127
Chapter 14. Impact of Changes in Antibiotic Use Practices on Nosocomial Infections and
Antimicrobial Resistance – Clostridium Difficile and Vancomycin-resistant Enterococcus
(VRE)....................................................................................................................................141
Chapter 15. Prevention of Nosocomial Urinary Tract Infections........................................149
Subchapter 15.1. Use of Silver Alloy Urinary Catheters .................................................149
Subchapter 15.2. Use of Suprapubic Catheters................................................................155
Chapter 16. Prevention of Intravascular Catheter-Associated Infections............................163
Subchapter 16.1. Use of Maximum Barrier Precautions During Central Venous Catheter
Insertion.............................................................................................................................165
Subchapter 16.2. Use of Central Venous Catheters Coated with Antibacterial or Antiseptic
Agents ...............................................................................................................................169
Subchapter 16.3. Use of Chlorhexidine Gluconate at the Central Venous Catheter
Insertion Site .....................................................................................................................175
Subchapter 16.4. Other Practices .....................................................................................178
vii

Chapter 17. Prevention of Ventilator-Associated Pneumonia.............................................185
Subchapter 17.1. Patient Positioning: Semi-recumbent Positioning and Continuous
Oscillation .........................................................................................................................185
Subchapter 17.2. Continuous Aspiration of Subglottic Secretions ..................................190
Subchapter 17.3. Selective Digestive Tract Decontamination.........................................193
Subchapter 17.4. Sucralfate and Prevention of VAP .......................................................198
Section C. Surgery, Anesthesia, and Perioperative Medicine ................................................204
Chapter 18. Localizing Care to High-Volume Centers .......................................................205
Chapter 19. Learning Curves for New Procedures – the Case of Laparoscopic
Cholecystectomy ..................................................................................................................213
Chapter 20. Prevention of Surgical Site Infections ..............................................................221
Subchapter 20.1. Prophylactic Antibiotics........................................................................221
Subchapter 20.2. Perioperative Normothermia................................................................231
Subchapter 20.3. Supplemental Perioperative Oxygen.....................................................234
Subchapter 20.4. Perioperative Glucose Control ..............................................................237
Chapter 21. Ultrasound Guidance of Central Vein Catheterization ....................................245
Chapter 22. The Retained Surgical Sponge.........................................................................255
Chapter 23. Pre-Anesthesia Checklists To Improve Patient Safety ....................................259
Chapter 24. The Impact Of Intraoperative Monitoring On Patient Safety ..........................265
Chapter 25. Beta-blockers and Reduction of Perioperative Cardiac Events .......................271
Section D. Safety Practices for Hospitalized or Institutionalized Elders................................280
Chapter 26. Prevention of Falls in Hospitalized and Institutionalized Older People........281
Subchapter 26.1. Identification Bracelets for High-Risk Patients ...................................284
Subchapter 26.2. Interventions that Decrease the Use of Physical Restraints .................287
Subchapter 26.3. Bed Alarms...........................................................................................291
Subchapter 26.4. Special Hospital Flooring Materials to Reduce Injuries from Patient Falls
...........................................................................................................................................293
Subchapter 26.5. Hip Protectors to Prevent Hip Fracture ................................................295
Chapter 27. Prevention of Pressure Ulcers in Older Patients ..............................................301
Chapter 28. Prevention of Delirium in Older Hospitalized Patients ...................................307
Chapter 29. Multidisciplinary Geriatric Consultation Services...........................................313
Chapter 30. Geriatric Evaluation and Management Units for Hospitalized Patients ...........323
Section E. General Clinical Topics.........................................................................................331
Chapter 31. Prevention of Venous Thromboembolism .......................................................333
Chapter 32. Prevention of Contrast-Induced Nephropathy .................................................349
Chapter 33. Nutritional Support ..........................................................................................359
Chapter 34. Prevention of Clinically Significant Gastrointestinal Bleeding in Intensive Care
Unit Patients .........................................................................................................................368
Chapter 35. Reducing Errors in the Interpretation of Plain Radiographs and Computed
Tomography Scans ...............................................................................................................376
Chapter 36. Pneumococcal Vaccination Prior to Hospital Discharge .................................385
36.1. Vaccine Effectiveness .............................................................................................386
36.2. Vaccine Delivery Methods......................................................................................388
viii

Chapter 37. Pain Management.............................................................................................396
Subchapter 37.1. Use of Analgesics in the Acute Abdomen ...........................................396
Subchapter 37.2. Acute Pain Services..............................................................................400
Subchapter 37.3. Prophylactic Antiemetics During Patient-controlled Analgesia Therapy
...........................................................................................................................................404
Subchapter 37.4. Non-pharmacologic Interventions for Postoperative Plan ...................406
Section F. Organization, Structure, and Culture ....................................................................411
Chapter 38. “Closed” Intensive Care Units and Other Models of Care for Critically Ill
Patients..................................................................................................................................413
Chapter 39. Nurse Staffing, Models of Care Delivery, and Interventions ..........................423
Chapter 40. Promoting a Culture of Safety ..........................................................................447
Section G. Systems Issues and Human Factors ......................................................................458
Chapter 41. Human Factors and Medical Devices ..............................................................459
Subchapter 41.1. The Use of Human Factors in Reducing Device-related Medical Errors
...........................................................................................................................................459
Subchapter 41.2. Refining the Performance of Medical Device Alarms .........................462
Subchapter 41.3. Equipment Checklists in Anesthesia....................................................466
Chapter 42. Information Transfer.........................................................................................471
Subchapter 42.1. Information Transfer Between Inpatient and Outpatient Pharmacies..471
Subchapter 42.2. Sign-Out Systems for Cross-Coverage ................................................476
Subchapter 42.3. Discharge Summaries and Follow-up ..................................................479
Subchapter 42.4. Notifying Patients of Abnormal Results ..............................................482
Chapter 43. Prevention of Misidentifications......................................................................487
Subchapter 43.1. Bar Coding ...........................................................................................487
Subchapter 43.2. Strategies to Avoid Wrong-Site Surgery..............................................494
Chapter 44. Crew Resource Management and its Applications in Medicine......................501
Chapter 45. Simulator-Based Training and Patient Safety..................................................511
Chapter 46. Fatigue, Sleepiness, and Medical Errors..........................................................519
Chapter 47. Safety During Transport of Critically Ill Patients............................................534
Subchapter 47.1. Interhospital Transport ..........................................................................536
Subchapter 47.2. Intrahospital Transport .........................................................................538
Section H. Role of the Patient.................................................................................................544
Chapter 48. Procedures For Obtaining Informed Consent ..................................................546
Chapter 49. Advance Planning For End-of-Life Care.........................................................556
Chapter 50. Other Practices Related to Patient Participation ..............................................569
PART IV. PROMOTING AND IMPLEMENTING SAFETY PRACTICES .......................................573
Chapter 51. Practice Guidelines ..........................................................................................575
Chapter 52. Critical Pathways .............................................................................................581
Chapter 53. Clinical Decision Support Systems..................................................................589
Chapter 54. Educational Techniques Used in Changing Provider Behavior.......................595
Chapter 55. Legislation, Accreditation, and Market-Driven and Other Approaches to
Improving Patient Safety......................................................................................................601
ix

PART V. ANALYZING THE PRACTICES .............................................................................611
Chapter 56. Methodology for Summarizing the Evidence for the Practices .......................613
Chapter 57. Practices Rated by Strength of Evidence .........................................................619
Chapter 58. Practices Rated by Research Priority...............................................................627
Chapter 59. Listing of All Practices, Categorical Ratings, and Comments........................633
Appendix: List of Contributors .................................................................................649
Index............................................................................................................................655
x

Summary
Overview
Patient safety has become a major concern of the general public and of policymakers at
the State and Federal levels. This interest has been fueled, in part, by news coverage of
individuals who were the victims of serious medical errors and by the publication in 1999 of the
Institute of Medicine’s (IOM’s) report To Err is Human: Building a Safer Health System. In its
report, IOM highlighted the risks of medical care in the United States and shocked the
sensibilities of many Americans, in large part through its estimates of the magnitude of medicalerrors-related
deaths (44,000 to 98,000 deaths per year) and other serious adverse events. The
report prompted a number of legislative and regulatory initiatives designed to document errors
and begin the search for solutions. But Americans, who now wondered whether their next
doctor’s or hospital visit might harm rather than help them, began to demand concerted action.
Three months after publication of the IOM report, an interagency Federal government
group, the Quality Interagency Coordination Task Force (QuIC), released its response, Doing
What Counts for Patient Safety: Federal Actions to Reduce Medical Errors and Their Impact.
That report, prepared at the President’s request, both inventoried on-going Federal actions to
reduce medical errors and listed more than 100 action items to be undertaken by Federal
agencies.
An action promised by the Agency for Healthcare Research and Quality (AHRQ), the
Federal agency leading efforts to research and promote patient safety, was “the development and
dissemination of evidence-based, best safety practices to provider organizations.” To initiate the
work to be done in fulfilling this promise, AHRQ commissioned the University of California at
San Francisco (UCSF) – Stanford University Evidence-based Practice Center (EPC) in January
2001 to review the scientific literature regarding safety improvement. To accomplish this, the
EPC established an Editorial Board that oversaw development of this report by teams of content
experts who served as authors.
Defining Patient Safety Practices
Working closely with AHRQ and the National Forum for Quality Measurement and
Reporting (the National Quality Forum, or NQF)—a public–private partnership formed in 1999
to promote a national health care quality agenda the EPC began its work by defining a patient
safety practice as a type of process or structure whose application reduces the probability of
adverse events resulting from exposure to the health care system across a range of diseases and
procedures.
This definition is consistent with the dominant conceptual framework in patient safety,
which holds that systemic change will be far more productive in reducing medical errors than
will targeting and punishing individual providers. The definition’s focus on actions that cut
across diseases and procedures also allowed the research team to distinguish patient safety
activities from the more targeted quality improvement practices (e.g., practices designed to
increase the use of beta-blockers in patients who are admitted to the hospital after having a
myocardial infarction). The editors recognize, however, that this distinction is imprecise.
1

This evidence-based review also focuses on hospital care as a starting point because the
risks associated with hospitalization are significant, the strategies for improvement are better
documented there than in other health care settings, and the importance of patient trust is
paramount. The report, however, also considers evidence regarding other sites of care, such as
nursing homes, ambulatory care, and patient self-management.
The results of this EPC study will be used by the NQF to identify a set of proven patient
safety practices that should be used by hospitals. Identification of these practices by NQF will
allow patients throughout the nation to evaluate the actions their hospitals and/or health care
facilities have taken to improve safety.
Reporting The Evidence
As is typical for evidence-based reviews, the goal was to provide a critical appraisal of
the evidence on the topic. This information would then be available to others to ensure that no
practice unsupported by evidence would be endorsed and that no practice substantiated by a high
level of proof would lack endorsement. Readers familiar with the state of the evidence regarding
quality improvement in areas of health care where this has been a research priority (e.g.,
cardiovascular care) may be surprised and even disappointed, by the paucity of high quality
evidence in other areas of health care for many patient safety practices. One reason for this is
the relative youth of the field. Just as there had been little public recognition of the risks of
health care prior to the first IOM report, there has been relatively little attention paid to such
risks – and strategies to mitigate them – among health professionals and researchers.
Moreover, there are a number of methodologic reasons why research in patient safety is
particularly challenging. Many practices (e.g., the presence of computerized physician order
entry systems, modifying nurse staffing levels) cannot be the subject of double-blind studies
because their use is evident to the participants. Second, capturing all relevant outcomes,
including “near misses” (such as a nurse catching an excessive dosage of a drug just before it is
administered to a patient) and actual harm, is often very difficult. Third, many effective practices
are multidimensional, and sorting out precisely which part of the intervention works is often
quite challenging. Fourth, many of the patient safety problems that generate the most concern
(wrong-site surgery, for example) are uncommon enough that demonstrating the success of a
“safety practice” in a statistically meaningful manner with respect to outcomes is all but
impossible.
Finally, establishing firm epidemiologic links between presumed (and accepted) causes
and adverse events is critical, and frequently difficult. For instance, in studying an intuitively
plausible “risk factor” for errors, such as “fatigue,” analyses of errors commonly reveal the
presence of fatigued providers (because many health care providers work long hours and/or late
at night). The question is whether or not fatigue is over-represented among situations that lead
to errors. The point is not that the problem of long work-hours should be ignored, but rather that
strong epidemiologic methods need to be applied before concluding that an intuitive cause of
errors is, in fact, causal.
Researchers now believe that most medical errors cannot be prevented by perfecting the
technical work of individual doctors, nurses, or pharmacists. Improving patient safety often
involves the coordinated efforts of multiple members of the health care team, who may adopt
strategies from outside health care. The report reviews several practices whose evidence came
2

from the domains of commercial aviation, nuclear safety, and aerospace, and the disciplines of
human factors engineering and organizational theory. Such practices include root cause analysis,
computerized physician order entry and decision support, automated medication dispensing
systems, bar coding technology, aviation-style preoperative checklists, promoting a “culture of
safety,” crew resource management, the use of simulators in training, and integrating human
factors theory into the design of medical devices and alarms. In reviewing these practices, the
research team sought to be flexible regarding standards of evidence, and included research
evidence that would not have been considered for medical interventions. For example, the
randomized trial that is appropriately hailed as the “gold standard” in clinical medicine is not
used in aviation, as this design would not capture all relevant information. Instead, detailed case
studies and industrial engineering research approaches are utilized.
Methodology
To facilitate identification and evaluation of potential patient safety practices, the
Editorial Board divided the content for the project into different domains. Some cover “content
areas,” including traditional clinical areas such as adverse drug events, nosocomial infections,
and complications of surgery, but also less traditional areas such as fatigue and information
transfer. Other domains consist of practices drawn from broad (primarily nonmedical)
disciplines likely to contain promising approaches to improving patient safety (e.g., information
technology, human factors research, organizational theory). Once this list was created—with
significant input from patient safety experts, clinician–researchers, AHRQ, and the NQF Safe
Practices Committee—the editors selected teams of authors with expertise in the relevant subject
matter and/or familiarity with the techniques of evidence-based review and technology appraisal.
The authors were given explicit instructions regarding search strategies for identifying
safety practices for evaluation (including explicit inclusion and exclusion criteria) and criteria for
assessing each practice’s level of evidence for efficacy or effectiveness in terms of study design
and study outcomes. Some safety practices did not meet the inclusion criteria because of the
paucity of evidence regarding efficacy or effectiveness but were included in the report because
an informed reader might reasonably expect them to be evaluated or because of the depth of
public and professional interest in them. For such high profile topics (such as bar coding to
prevent misidentifications), the researchers tried to fairly present the practice’s background, the
experience with the practice thus far, and the evidence (and gaps in the evidence) regarding the
practice’s value.
For each practice, authors were instructed to research the literature for information on:
• prevalence of the problem targeted by the practice;
• severity of the problem targeted by the practice;
• the current utilization of the practice;
• evidence on efficacy and/or effectiveness of the practice;
• the practice’s potential for harm;
• data on cost, if available; and
• implementation issues.
The report presents the salient elements of each included study (e.g., study design,
population/setting, intervention details, results), and highlights any important weaknesses and
3

iases of these studies. Authors were not asked to formally synthesize or combine the evidence
across studies (e.g., perform a meta-analysis) as part of their task.
The Editorial Board and the Advisory Panel reviewed the list of domains and practices to
identify gaps in coverage. Submitted chapters were reviewed by the Editorial Board and revised
by the authors, aided by feedback from the Advisory Panel. Once the content was finalized, the
editors analyzed and ranked the practices using a methodology summarized below.
Summarizing the Evidence and Rating the Practices
Because the report is essentially an anthology of a diverse and extensive group of patient
safety practices with highly variable relevant evidence, synthesizing the findings was
challenging, but necessary to help readers use the information. Two of the most obvious uses for
this report are: 1) to inform efforts of providers and health care organizations to improve the
safety of the care they provide, and 2) to inform AHRQ, other research agencies, and
foundations about potential fruitful investments for their research support. Other uses of the
information are likely. In fact, the National Quality Forum plans to use this report to help
identify a list of patient safety practices that consumers and others should know about as they
choose among the health care provider organizations to which they have access.
In an effort to assist both health care organizations interested in taking substantive
actions to improve patient safety and research funders seeking to spend scarce resources wisely,
AHRQ asked the EPC to rate the evidence and rank the practices by opportunity for safety
improvement and by research priority. This report, therefore, contains two lists.
To create these lists, the editors aimed to separate the practices that are most promising or
effective from those that are least so on a range of dimensions, without implying any ability to
calibrate a finely gradated scale for those practices in between. The editors also sought to
present the ratings in an organized, accessible way while highlighting the limitations inherent in
their rating schema. Proper metrics for more precise comparisons (e.g., cost-effectiveness
analysis) require more data than are currently available in the literature.
Three major categories of information were gathered to inform the rating exercise:
• Potential Impact of the Practice: based on prevalence and severity of the patient
safety target, and current utilization of the practice;
• Strength of the Evidence Supporting the Practice: including an assessment of the
relative weight of the evidence, effect size, and need for vigilance to reduce any
potential negative collateral effects of the practice; and
• Implementation: considering costs, logistical barriers, and policy issues.
For all of these data inputs into the practice ratings, the primary goal was to find the best
available evidence from publications and other sources. Because the literature has not been
previously organized with an eye toward addressing each of these areas, most of the estimates
could be improved with further research, and some are informed by only general and somewhat
speculative knowledge. In the summaries, the editors have attempted to highlight those
assessments made with limited data.
The four-person editorial team independently rated each of the 79 practices using general
scores (e.g., High, Medium, Low) for a number of dimensions, including those italicized in the
4

section above. The editorial team convened for 3 days in June, 2001 to compare scores, discuss
disparities, and come to consensus about ratings for each category.
In addition, each member of the team considered the totality of information on potential
impact and support for a practice to score each of these factors on a 0 to 10 scale (creating a
“Strength of the Evidence” list). For these ratings, the editors took the perspective of a leader of
a large health care enterprise (e.g., a hospital or integrated delivery system) and asked the
question, “If I wanted to improve patient safety at my institution over the next 3 years and
resources were not a significant consideration, how would I grade this practice?” For this
rating, the Editorial Board explicitly chose not to formally consider the difficulty or cost of
implementation in the rating. Rather, the rating simply reflected the strength of the evidence
regarding the effectiveness of the practice and the probable impact of its implementation on
reducing adverse events related to health care exposure. If the patient safety target was rated as
“High” impact and there was compelling evidence (i.e., “High” relative study strength) that a
particular practice could significantly reduce (e.g., “Robust” effect size) the negative
consequences of exposure to the health care system (e.g., hospital-acquired infections), raters
were likely to score the practice close to 10. If the studies were less convincing, the effect size
was less robust, or there was a need for a “Medium” or “High” degree of vigilance because of
potential harms, then the rating would be lower.
At the same time, the editors also rated the usefulness of conducting more research on
each practice, emphasizing whether there appeared to be questions that a research program might
have a reasonable chance of addressing successfully (creating a “Research Priority” list). Here,
they asked themselves, “If I were the leader of a large agency or foundation committed to
improving patient safety, and were considering allocating funds to promote additional research,
how would I grade this practice?” If there was a simple gap in the evidence that could be
addressed by a research study or if the practice was multifaceted and implementation could be
eased by determining the specific elements that were effective, then the research priority was
high. (For this reason, some practices are highly rated on both the “Strength of the Evidence”
and “Research Priority” lists.) If the area was one of high potential impact (i.e., large number of
patients at risk for morbid or mortal adverse events) and a practice had been inadequately
researched, then it would also receive a relatively high rating for research need. Practices might
receive low research scores if they held little promise (e.g., relatively few patients are affected by
the safety problem addressed by the practice or a significant body of knowledge already
demonstrates the practice’s lack of utility). Conversely, a practice that was clearly effective, low
cost, and easy to implement would not require further research and would also receive low
research scores.
In rating both the strength of the evidence and the research priority, the purpose was not
to report precise 0 to 10 scores, but to develop general “zones” or practice groupings. This is
important because better methods are available for making comparative ratings when the data
inputs are available. The relative paucity of the evidence dissuaded the editors from using a
more precise, sophisticated, but ultimately unfeasible, approach.
5

Clear Opportunities for Safety Improvement
The following 11 patient safety practices were the most highly rated (of the 79 practices
reviewed in detail in the full report and ranked in the Executive Summary Addendum, AHRQ
Publication No. 01-E057b) in terms of strength of the evidence supporting more widespread
implementation. Practices appear in descending order, with the most highly rated practices listed
first. Because of the imprecision of the ratings, the editors did not further divide the practices,
nor indicate where there were ties.
• Appropriate use of prophylaxis to prevent venous thromboembolism in
patients at risk;
• Use of perioperative beta-blockers in appropriate patients to prevent
perioperative morbidity and mortality;
• Use of maximum sterile barriers while placing central intravenous catheters
to prevent infections;
• Appropriate use of antibiotic prophylaxis in surgical patients to prevent
perioperative infections;
• Asking that patients recall and restate what they have been told during the
informed consent process;
• Continuous aspiration of subglottic secretions (CASS) to prevent ventilatorassociated
pneumonia;
• Use of pressure relieving bedding materials to prevent pressure ulcers;
• Use of real-time ultrasound guidance during central line insertion to prevent
complications;
• Patient self-management for warfarin (Coumadin) to achieve appropriate
outpatient anticoagulation and prevent complications;
• Appropriate provision of nutrition, with a particular emphasis on early enteral
nutrition in critically ill and surgical patients; and
• Use of antibiotic-impregnated central venous catheters to prevent catheterrelated
infections.
This list is generally weighted toward clinical rather than organizational matters, and
toward care of the very, rather than the mildly or chronically ill. Although more than a dozen
practices considered were general safety practices that have been the focus of patient safety
experts for decades (i.e., computerized physician order entry, simulators, creating a “culture of
safety,” crew resource management), most research on patient safety has focused on more
clinical areas. The potential application of practices drawn from outside health care has excited
the patient safety community, and many such practices have apparent validity. However, clinical
research has been promoted by the significant resources applied to it through Federal,
foundation, and industry support. Since this study went where the evidence took it, more clinical
practices rose to the top as potentially ready for implementation.
6

Clear Opportunities for Research
Until recently, patient safety research has had few champions, and even fewer champions
with resources to bring to bear. The recent initiatives from AHRQ and other funders are a
promising shift in this historical situation, and should yield important benefits.
In terms of the research agenda for patient safety, the following 12 practices rated most
highly, as follows:
• Improved perioperative glucose control to decrease perioperative
infections;
• Localizing specific surgeries and procedures to high volume centers;
• Use of supplemental perioperative oxygen to decrease perioperative
infections;
• Changes in nursing staffing to decrease overall hospital morbidity and
mortality;
• Use of silver alloy-coated urinary catheters to prevent urinary tract
infections;
• Computerized physician order entry with computerized decision support
systems to decrease medication errors and adverse events primarily due to
the drug ordering process;
• Limitations placed on antibiotic use to prevent hospital-acquired
infections due to antibiotic-resistant organisms;
• Appropriate use of antibiotic prophylaxis in surgical patients to prevent
perioperative infections;
• Appropriate use of prophylaxis to prevent venous thromboembolism in
patients at risk;
• Appropriate provision of nutrition, with a particular emphasis on early
enteral nutrition in critically ill and post-surgical patients;
• Use of analgesics in the patient with an acutely painful abdomen without
compromising diagnostic accuracy; and
• Improved handwashing compliance (via education/behavior change; sink
technology and placement; or the use of antimicrobial washing
substances).
Of course, the vast majority of the 79 practices covered in this report would benefit from
additional research. In particular, some practices with longstanding success outside of medicine
(e.g., promoting a culture of safety) deserve further analysis, but were not explicitly ranked due
to their unique nature and the present weakness of the evidentiary base in the health care
literature.
7

Conclusions
This report represents a first effort to approach the field of patient safety through the lens
of evidence-based medicine. Just as To Err is Human sounded a national alarm regarding patient
safety and catalyzed other important commentaries regarding this vital problem, this review
seeks to plant a seed for future implementation and research by organizing and evaluating the
relevant literature. Although all those involved tried hard to include all relevant practices and to
review all pertinent evidence, inevitably some of both were missed. Moreover, the effort to
grade and rank practices, many of which have only the beginnings of an evidentiary base, was
admittedly ambitious and challenging. It is hoped that this report provides a template for future
clinicians, researchers, and policy makers as they extend, and inevitably improve upon, this
work.
In the detailed reviews of the practices, the editors have tried to define (to the extent
possible from the literature) the associated costs—financial, operational, and political. However,
these considerations were not factored into the summary ratings, nor were judgments made
regarding the appropriate expenditures to improve safety. Such judgments, which involve
complex tradeoffs between public dollars and private ones, and between saving lives by
improving patient safety versus doing so by investing in other health care or non-health care
practices, will obviously be critical. However, the public reaction to the IOM report, and the
media and legislative responses that followed it, seem to indicate that Americans are highly
concerned about the risks of medical errors and would welcome public and private investment to
decrease them. It seems logical to infer that Americans value safety during a hospitalization just
as highly as safety during a transcontinental flight.
8

Evidence Report

PART I. OVERVIEW
Chapter 1. An Introduction to the Report
1.1. General Overview
1.2. How to Use this Report
1.3. Acknowledgments
Chapter 2. Drawing on Safety Practices from Outside Health Care
Chapter 3. Evidence-based Review Methodology
11

Chapter 1. An Introduction to the Report
1.1. General Overview
The Institute of Medicine’s (IOM) report, To Err is Human: Building a Safer Health
System, 1 highlighted the risks of medical care in the United States. Although its prose was
measured and its examples familiar to many in the health professions (for example, the studies
estimating that up to 98,000 Americans die each year from preventable medical errors were a
decade old), the report shocked the sensibilities of many Americans. More importantly, the
report undermined the fundamental trust that many previously had in the health care system.
The IOM report prompted a number of legislative and regulatory initiatives designed to
document errors and begin the search for solutions. These initiatives were further catalyzed by a
second IOM report entitled Crossing the Quality Chasm: A New Health System for the 21 st
Century, 2 which highlighted safety as one of the fundamental aims of an effective system. But
Americans, who now wondered whether their next health care encounter might harm rather than
help them, began to demand concerted action.
Making Health Care Safer represents an effort to determine what it is we might do in an
effort to improve the safety of patients. In January 2001, the Agency for Healthcare Research
and Quality (AHRQ), the Federal agency taking the lead in studying and promoting patient
safety, commissioned the UCSF-Stanford Evidence-based Practice Center (EPC) to review the
literature as it pertained to improving patient safety. In turn, the UCSF-Stanford EPC engaged 40
authors at 11 institutions around the United States to review more than 3000 pieces of literature
regarding patient safety practices. Although AHRQ expected that this evidence-based review
would have multiple audiences, the National Quality Forum (NQF)—a public-private partnership
formed in the Clinton Administration to promote a national quality agenda—was particularly
interested in the results as it began its task of recommending and implementing patient safety
practices supported by the evidence.
A Definition of “Patient Safety Practices”
One of our first tasks was to define “patient safety practices” in a manner that would
allow us and our reviewers to assess the relevant evidence. Given our task—producing a full
report in less than six months—a complete review of all practices associated with improving
health care quality was both impossible and off-point. Working closely with AHRQ and NQF,
we chose the following definition:
A Patient Safety Practice is a type of process or structure whose application reduces
the probability of adverse events resulting from exposure to the health care system
across a range of diseases and procedures.
A few elements of the definition deserve emphasis. First, our focus on processes and
structure allowed us to emphasize changing the system to make it safer rather than targeting and
removing individual “bad apples.” We recognize that when individuals repeatedly perform
poorly and are unresponsive to education and remediation, action is necessary. Nevertheless,
there is virtual unanimity among patient safety experts that a focus on systemic change will be
far more productive than an emphasis on finding and punishing poor performers.
Second, looking at crosscutting diseases and procedures allowed us to distinguish patient
safety activities from more targeted quality improvement practices. Admittedly, this
dichotomization is imprecise. All would agree that a practice that makes it less likely that a
13

patient will receive the wrong medication or have the wrong limb amputated is a patient safety
practice. Most would also agree that practices designed to increase the use of beta-blockers in
patients admitted to the hospital after myocardial infarction or to improve the technical
performance of hernia repair would be quality improvement strategies rather than patient safety
practices. When there was a close call, we generally chose to be inclusive. For example, we
included practices designed to increase the rate of appropriate prophylaxis against venous
thromboembolism, the appropriateness of pain management, and the ascertainment of patient
preferences regarding end-of-life care. We recognize that these practices blur the line somewhat
between safety and quality, but we believe that they are reasonable examples of ways to address
potential patient safety hazards.
Third, we realized it would be impossible to review every potential safety practice and
recognized that some gaps in the evidence were inevitable, so at times we reviewed illustrative
examples that might be broadly generalizable. For example:
• Methods to avoid misread radiographs (Chapter 35); where the content could
be relevant to analogous efforts to avoid misread electrocardiograms or
laboratory studies
• Decreasing the risk of dangerous drugs (Chapter 9), where the focus was on
anticoagulants, but similar considerations might be relevant for chemotherapy
and other high-risk drugs
• Localizing care to specialized providers reviews geriatric units and
intensivists (Chapters 30 and 38), but similar evidence may be relevant for the
rapidly growing field of hospitalists 3-5
• The use of ultrasound guidance for central line placement (Chapter 21); the
premise (decreasing the risk of an invasive procedure through radiologic
localization) may also be relevant for ultrasound guidance while performing
other challenging procedures, such as thoracentesis
A Focus on Hospital Care
Most of the literature regarding medical errors has been drawn from hospital care. 6-22 For
example, the two seminal studies on medical error 22,23 from which the oft-cited extrapolations of
yearly deaths from medical error were derived, have highlighted the risks of inpatient care. We
applaud recent studies examining the risks of errors in the ambulatory setting 24 but believe that
the hospital is an appropriate initial focus for an evidence-based review because the risks
associated with hospitalization are high, strategies for improvement are better documented, and
the importance of patient trust is paramount.
That said, the reader will see that we allowed the evidence to take us to other sites of
care. For example, although much of the literature regarding the occurrence and prevention of
adverse drug events is hospital-based, more recent literature highlights outpatient issues and is
included in this review. An example is the chapter on decreasing the risk of anticoagulant
treatment (Chapter 9), in which two of the most promising practices involve outpatient
anticoagulation clinics and patient self-monitoring at home. Similarly, strategies to prevent falls
or pressure ulcers are relevant to nursing home patients as well as those in hospitals, and many
studies that shed light on these issues come from the former setting.
14

The Evidence
Chapter 3 describes our strategy for evidence review. As in other evidence-based
reviews, we set the bar high. One would not want to endorse a practice unsupported by evidence,
nor withhold one substantiated by a high level of proof. In the end, we aimed to identify
practices whose supporting evidence was so robust that immediate widespread implementation
would lead to major improvements in patient safety. Additionally, we hoped to identify several
practices whose promise merited a considerable investment in additional research, but whose
evidentiary base was insufficient for immediate endorsement. The results of this effort are
summarized in Part V of the Report.
Readers familiar with the state of the evidence regarding quality improvement in areas
where this has been a research priority (eg, cardiovascular care) may be surprised and even
disappointed by the paucity of high quality evidence for many patient safety practices. The field
is young. Just as there had been little public recognition of the risks of health care prior to the
first IOM report, there has been relatively little attention paid to such risks—and strategies to
mitigate them—among health professionals and researchers. Nevertheless, we found a number of
practices supported by high quality evidence for which widespread implementation would save
many thousands of lives.
Moreover, there are important methodologic reasons why research in patient safety is
particularly challenging. First is the problem of blinding. The physician who has begun to use a
new computerized order entry system cannot be blinded to the intervention or its purpose.
Second, it is sometimes difficult to measure important outcomes. As in aviation, enormous
benefits can be reaped from analyzing “near misses” (with no ultimate harm to patients), 25,26 and
yet these outcomes cannot be reliably counted in the absence of potentially obtrusive, and often
very expensive observation. Third, many effective practices are multidimensional, and sorting
out precisely which part of the intervention works is often quite challenging. Fourth, many of the
patient safety problems that generate the most concern (wrong-site surgery, for example) are
probably uncommon. This makes demonstrating the success of a “safety practice” in a
statistically meaningful manner with respect to outcomes all but impossible.
Finally, establishing firm epidemiologic links between presumed (and accepted) causes
and adverse events is critical, and frequently difficult. For instance, verbal orders from doctors to
nurses are regarded as a cause of medication errors almost as matter of dogma, with many
hospitals prohibiting or strongly discouraging this practice except in emergency situations. 27 Yet,
the one study that we could identify that specifically and comprehensively addressed this issue 28
actually reported fewer errors among verbal medication orders compared with written medication
orders. A similar relationship might be found studying other intuitively plausible “risk factors”
for errors, such as “fatigue.” Because many health care providers work long hours and/or late at
night, analyses of errors will commonly reveal fatigued providers. The question is whether or not
fatigue is over-represented among situations that lead to errors. As discussed in Chapter 46, the
evidence supporting fatigue as a contributor to adverse events is surprisingly mixed. The point is
not that the problem of long work-hours should be ignored, but rather that strong epidemiologic
methods need to be applied before concluding that an intuitive cause of errors is in fact causal.
These methodologic issues are further explored in Chapters 3 (methods for analyzing the
individual practices) and 56 (methods for summarizing the overall evidence).
Improving patient safety is a team effort, and the playbook is often drawn from fields
outside of health care. Most medical errors cannot be prevented by perfecting the technical work
of individual doctors, nurses or pharmacists. Improving patient safety often involves the
15

coordinated efforts of multiple members of the health care team, who may adopt strategies from
outside health care. Thus, our teams of authors and advisors included physicians, pharmacists,
nurses, and experts from non-medical fields. The literature we reviewed was often drawn from
journals, books, or Web sites that will not be on most doctors’ reading lists. We reviewed several
promising practices whose evidence came from the domains of commercial aviation, nuclear
safety, and aerospace, and the disciplines of human factors engineering and organizational
theory. In reviewing these practices, we tried to be flexible regarding standards of evidence. For
example, the randomized trial that is appropriately hailed as the “gold standard” in health care is
rarely used in aviation, which instead relies on analyses of detailed case studies and industrial
engineering research approaches. (Examples and additional discussion of this issue can be found
in Chapter 2.)
We also limited our discussion to the existing practices, recognizing that future
technology may make the ones we reviewed obsolete. For example, much of the struggle to find
safe ways to administer warfarin (Chapter 9) would be rendered moot by the development of a
much safer, but equally effective oral anticoagulant that did not require monitoring. Similarly,
the evidence regarding changing the flooring of rooms to decrease falls (Subchapter 26.4)
indicated that present options may decrease the harm from falls but actually increase their rate.
Clearly, a better surface would make falls both less likely and less harmful. Such a surface has
not yet been tested.
Finally, we have tried to define (to the extent possible from the literature) the costs—
financial, operational, and political—associated with the patient safety practices we considered.
However, we have not made judgments regarding the appropriate expenditures to improve
safety. These judgments, which involve complex tradeoffs between public dollars and private
ones, and between saving lives by improving patient safety versus doing so by investing in other
health care or non-health care practices, will obviously be critical. However, the public reaction
to the IOM report, and the media and legislative responses that followed it, seem to indicate that
Americans are highly concerned about the risks of medical errors and would welcome public and
private investment to decrease them. It seems logical to infer that Americans value safety during
a hospitalization just as highly as safety during a transcontinental flight.
The Decision to Include and Exclude Practices
The patient safety/quality interface was only one of several areas that called for
judgments regarding which practices to include or exclude from the Report. In general (and quite
naturally for an evidence-based review), we excluded those practices for which we found little or
no supporting evidence. However, we recognize that patient safety is of great public and
professional interest, and that the informed reader might expect to find certain topics in such a
review. Therefore, we included several areas notwithstanding their relatively meager evidentiary
base. For such high profile topics (such as bar coding to prevent misidentifications, Subchapter
43.1), we tried to fairly present the practice’s background, the experience with the practice thus
far, and the evidence (and gaps in the evidence) regarding its value. In many of these cases, we
end by encouraging additional study or demonstration projects designed to prove whether the
practices live up to their promise.
Conversely, another very different group of practices lacked evidence and were excluded
from the review. These practices were characterized by their largely self-evident value (in
epidemiologic terms, their “face validity”). For example, large randomized studies of the
removal of concentrated potassium chloride from patient care floors surely are not necessary in
order to recommend this practice as a sensible way of preventing egregious errors that should
16

never occur. Although some of these types of practices were not included in this “evidencebased”
Report, the reader should not infer their exclusion as a lack of endorsement.
A cautionary note is in order when considering such “obviously beneficial” practices.
Even an apparently straightforward practice like “signing the site” to prevent surgery or
amputation of the wrong body part may lead to unexpected opportunities for error. As mentioned
in Subchapter 43.2, some surgeons adopt the practice of marking the intended site, while others
mark the site to avoid. The clinical research literature furnishes enough examples of practices
that everyone “knew” to be beneficial but proved not to be (or even proved to be harmful) once
good studies were conducted (antiarrhythmic therapy for ventricular ectopy 29 or hormone
replacement therapy to prevent cardiac deaths, 30 for example) that it is reasonable to ask for
high-quality evidence for most practices. This is particularly true when practices are expensive,
complex to implement, or carry their own risks.
Other Content Issues
There may appear to some readers to be an inordinate focus on clinical issues versus
more general patient safety practices. In this and other matters, we went where the evidence took
us. Although more than a dozen chapters of the Report consider general safety practices that
have been the focus of many patient safety experts for decades (ie, computerized order entry,
simulators, crew resource management), most research on patient safety, in fact, has focused on
more clinical matters. It is likely that some of this is explained by the previous “disconnect”
between research in patient safety and its application. We are hopeful that the Report helps to
bridge this gap. We also think it likely that clinical research has been promoted by the significant
resources applied to it through Federal, foundation, and industry support. Until recently, patient
safety research has had few champions, and even fewer champions with resources. The recent
initiatives from AHRQ and other funders are a promising shift in this historical situation, and
should yield important benefits.
The reader will notice that there is relatively little specific coverage of issues in
pediatrics, obstetrics, and psychiatry. Most of the patient safety practices we reviewed have
broad applicability to those fields as well as larger fields such as surgery and medicine. Much of
the research in the former fields was too disease-specific to include in this volume. For example,
practices to improve the safety of childbirth, although exceptionally important, were excluded
because they focused on the care of patients with a single “condition,” just as we excluded
research focused specifically on the care of patients with pneumonia or stroke.
Readers may also be surprised by the relatively small portion of the Report devoted to the
prevention of high-profile and “newsworthy” errors. Even if much of the national attention to
patient safety stemmed from concerns about wrong-site surgery or transfusion mix-ups, in fact
these are not the dominant patient safety problems today. If widespread use of hip protectors
(Subchapter 26.5) leads to a marked decrease in injuries from patient falls, implementing this
safety practice would be more important than preventing the few wrong-site surgeries each year,
although the former seem far less likely to garner attention in a tabloid.
Conclusions
Making Health Care Safer represents a first effort to approach the field of patient safety
through the lens of evidence-based medicine. Just as To Err is Human sounded a national alarm
regarding patient safety and catalyzed other important commentaries regarding this vital
problem, this review is a germinal effort to mine the relevant literature. Although we and the
authors tried hard to include all relevant practices and to review all pertinent evidence, we
17

inevitably missed some of both. Moreover, our effort to rank practices (Part V), many of which
have only the beginnings of an evidentiary base, was admittedly ambitious and challenging. We
hope that the Report provides a template for future clinicians, researchers, and policy makers as
they extend, and inevitably improve upon, our work.
1.2. How to Use this Report
Organizational Framework
This document is divided into five parts:
Part I – The overview introduces many of the methodologic, content, and policy issues.
Part II – We describe, and present the evidence regarding 2 practices that are used to report and
respond to patient safety problems: incident reporting and root cause analysis. Since both these
“practices” have relevance for all of the patient safety targets and practices covered in Part III,
we neither grade them nor rank them.
Part III – In 45 chapters, we review the evidence regarding the utility of 79 patient safety
practices. Each chapter is structured in a standard fashion, as follows:
• Background – of the patient safety problem and the practice;
• Practice Description – in which we try to present the practice at a level of
detail that would allow a reader to determine the practice’s applicability to
their setting;
• Prevalence and Severity of the Target Safety Problem – Here, we try to
answer the following questions: How common is the safety problem the
practice is meant to address? How often does the problem lead to harm? How
bad is the harm when it occurs?;
• Opportunities for Impact – In this section, we consider the present-day use of
the patient safety practice. For example, we found that the use of “unit-dose”
drug dispensing was quite common in US hospitals, and thus the opportunity
to make an impact with wider dissemination of this practice was relatively
low. Conversely, computerized physician order entry is still relatively
uncommon, and therefore (assuming it is effective), its widespread
implementation could have a far larger impact;
• Study Designs – We review the designs of the major studies evaluating the
practice. Similarly, Study Outcomes looks at the kinds of outcomes (eg,
adverse drug events, surgical complications, mortality) that were considered.
Our criteria for grading the evidence related to both design and outcomes
(more information on other methodologic issues appears in Chapter 3);
• Evidence for Effectiveness of the Practice – Here, the authors summarize the
findings of the studies and comment on any methodologic concerns that might
effect the strength of these findings. This section is often accompanied by
tables summarizing the studies and their findings;
18

• Potential for Harm – Many practices that are effective in improving patient
safety nonetheless carry the potential for harm. More widespread use of
antibiotic prophylaxis or antibiotic-impregnated urinary or vascular catheters
could prevent individual hospital-acquired infections yet breed antibiotic
resistance. Increasing the use of barrier precautions could also prevent
infections, but might lead caregivers to visit patients less often. These sections
do not imply that harm is inevitable; rather they highlight the issues that
require vigilance during the implementation of effective practices;
• Costs and Implementation – Here we consider the costs and other challenges
of implementing the practice. We tried to uncover data related to the true costs
of implementation (How much does an automatic drug dispensing machine
cost a pharmacy?), but also considered some of the potential offsets when
there were data available. We also considered issues of feasibility: How much
behavior change would be necessary to implement the practice? Would there
be major political concerns or important shifts in who pays for care or is
compensated for providing it? We tried not to assign values to such issues, but
rather to present them so that policy makers could consider them; and
• Comment – Here, the authors highlight the state of the evidence, elucidate key
implementation issues, and define a potential research agenda.
Part IV – In many ways a mirror of Part II, Part IV considers the ways in which patient safety
practices can be implemented. The evidence is reviewed, and some of the benefits and
limitations of various strategies are analyzed. As with Part II, we neither grade nor rank these
“practices” in Part V since each of these strategies can be applied to most of the patient safety
targets and practices covered in Part III.
Part V – Here we analyze the practices. Using methods described in Chapter 56, we synthesize
the evidence in Part III to grade and rank the patient safety practices across two major
dimensions:
• Does the evidence support implementation of the practice to improve patient safety?
• Does the evidence support additional research into the practice?
Tips for Users of the Report
We envision that this evidence-based report of patient safety practices will be useful to a
wide audience.
Policy makers may use its contents and recommendations to promote or fund the
implementation of certain practices. Similarly, local health care organization leaders (including
leaders of hospitals, medical groups, or integrated delivery systems) may use the data and
analysis to choose which practices to consider implementing or further promoting at their
institutions.
Researchers will identify a wealth of potential research opportunities. This document is,
in many ways, a road map for future research into patient safety. Those who fund research,
including (but not limited to) AHRQ, which sponsored this report, will find literally dozens of
areas ripe for future studies. In some cases, such studies may be expensive randomized
controlled trials, while other practices may require a simple meta-analysis or cost-effectiveness
analysis to tip the scales toward or away from recommending a practice.
19

Clinicians and trainees will, we hope, find the material both interesting and relevant to
their practices. One of the salutary consequences of the IOM’s reports has been their impact on
the attitudes of our future health care providers. We have noticed at our institutions that students
and post-graduate trainees in medicine, nursing, and pharmacy are increasingly taking a systems
approach to health care. Several of us have heard medical residents refer to issues as “patient
safety problems” that beg for a “systems solution” over the past two years, terms that were
absent from the medical ward a few years earlier. Clinicians must be part of the solutions to
patient safety problems, and their increasing interest in the field is an exceedingly hopeful sign.
Finally, although not primarily written for patients and their families, we recognize the
broad public interest in, and concern about patient safety and believe that much of the material
will be compelling and potentially useful to the public. For years quality advocates have
lamented the relatively small impact that “quality report cards” appear to have on patients’
choices of health care providers and institutions. One study demonstrated that patients were more
likely to respond to a newspaper report of an egregious error than such quality report cards. 31
These data indicate that patients may be interested in knowing whether their institutions,
providers, and health plans are proactive in implementing practices that demonstrably decrease
the risk of adverse events. Also, any general reader is likely to come away from this Report with
heightened sensitivity to the unique challenges that the health care industry—which aims to
provide compassionate, individualized care in a dynamic, organizationally and politically
complex, and technologically fluid environment—faces in improving safety, and the significant
strides that have already been made. Continued improvement will require the infusion of
substantial resources, and the public debate about their source, quantity, and target is likely to be
lively and very important.
1.3. Acknowledgments
We are grateful to our authors, whose passion for the evidence overcame the tremendous
time pressure driven by an unforgiving timeline. They succumbed to our endless queries, as
together we attempted to link the literature from a variety of areas into an evidence-based
repository whose presentation respected an overarching framework, notwithstanding the variety
of practices reviewed. Our Managing Editor, Amy J. Markowitz, JD, edited the manuscript
with incredible skill and creativity. Each part was reviewed and improved multiple times because
of her unending dedication to the project. Susan B. Nguyen devoted countless hours assisting
the team in every imaginable task, and was critical in keeping the project organized and on track.
Invaluable editorial and copyediting assistance was provided by Kathleen Kerr, Talia Baruth,
and Mary Whitney. We thank Phil Tiso for helping produce the electronic version of the
Report. We also thank A. Eugene Washington, MD, MSc, Director of the UCSF-Stanford
Evidence-based Practice Center, for his guidance and support.
We were aided by a blue-ribbon Advisory Panel whose suggestions of topics, resources,
and ideas were immensely helpful to us throughout the life of the project. Given the time
constraints, there were doubtless many constructive suggestions that we were unable to
incorporate into the Report. The responsibility for such omissions is ours alone. We want to
recognize the important contribution of our advisors to our thinking and work. The Advisory
Panel’s members are:
• David M. Gaba, MD, Director, Patient Safety Center of Inquiry at Veterans
Affairs Palo Alto Health Care System, Professor of Anesthesia, Stanford
University School of Medicine
20

• John W. Gosbee, MD, MS, Director of Patient Safety Information Systems,
Department of Veterans Affairs National Center for Patient Safety
• Peter V. Lee, JD, President and Chief Executive Officer, Pacific Business
Group on Health
• Arnold Milstein, MD, MPH, Medical Director, Pacific Business Group on
Health; National Health Care Thought Leader, William M. Mercer, Inc.
• Karlene H. Roberts, PhD, Professor, Walter A. Haas School of Business,
University of California, Berkeley
• Stephen M. Shortell, PhD, Blue Cross of California Distinguished Professor
of Health Policy and Management and Professor of Organization Behavior,
University of California, Berkeley
We thank the members of the National Quality Forum’s (NQF) Safe Practices
Committee, whose co-chairs, Maureen Bisognano and Henri Manasse, Jr., PhD, ScD, and
members helped set our course at the outset of the project. NQF President and Chief Executive
Officer Kenneth W. Kizer, MD, MPH, Vice President Elaine J. Power, and Program Director
Laura N. Blum supported the project and coordinated their committee members’ helpful
suggestions as the project proceeded. We appreciate their patience during the intensive research
period as they waited for the final product.
Finally, we are indebted to the Agency for Healthcare Research and Quality (AHRQ),
which funded the project and provided us the intellectual support and resources needed to see it
to successful completion. Jacqueline Besteman, JD of AHRQ’s Center for Practice and
Technology Assessment, and Gregg Meyer, MD, and Nancy Foster, both from AHRQ’s Center
for Quality Improvement and Patient Safety, were especially important to this project. John
Eisenberg, MD, MBA, AHRQ’s Administrator, was both a tremendous supporter of our work
and an inspiration for it.
The Editors
San Francisco and Palo Alto, California
June, 2001
References
1. Kohn L, Corrigan J, Donaldson M, editors. To Err Is Human: Building a Safer Health
System. Washington, DC: Committee on Quality of Health Care in America, Institute of
Medicine. National Academy Press; 2000.
2. Committee on Quality of Health Care in America, Institute of Medicine, editor. Crossing
the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National
Academy Press; 2001.
3. Wachter RM, Goldman L. The emerging role of "hospitalists" in the American health care
system. N Engl J Med. 1996;335:514-517.
4. Wachter RM, Katz P, Showstack J, Bindman AB, Goldman L. Reorganizing an academic
medical service: impact on cost, quality, patient satisfaction, and education. JAMA.
1998;279:1560-1565.
21

5. Meltzer DO, Manning WG, Shah M, Morrison J, Jin L, Levinson W. Decreased length of
stay, costs and mortality in a randomized trial of academic hospitalists. Paper presented at:
Society of General Internal Medicine-24th Annual Meeting; May 2-5, 2001; San Diego,
CA.
6. Safren MA, Chapanis A. A critical incident study of hospital medication errors. Hospitals.
1960;34:32-34.
7. Mills DH, editor. Report on the medical insurance feasibility study / sponsored jointly by
California Medical Association and California Hospital Association. San Francisco: Sutter
Publications, Inc.; 1977.
8. Bosk CL. Forgive and remember : managing medical failure. Chicago: University of
Chicago Press; 1979.
9. Abramson NS, Wald KS, Grenvik AN, Robinson D, Snyder JV. Adverse occurrences in
intensive care units. JAMA. 1980;244:1582-1584.
10. Barker KN, Mikeal RL, Pearson RE, Illig NA, Morse ML. Medication errors in nursing
homes and small hospitals. Am J Hosp Pharm. 1982;39:987-991.
11. Cooper JB, Newbower RS, Kitz RJ. An analysis of major errors and equipment failures in
anesthesia management: considerations for prevention and detection. Anesthesiology.
1984;60:34-42.
12. Rubins HB, Moskowitz MA. Complications of care in a medical intensive care unit. J Gen
Intern Med. 1990;5:104-109.
13. Silber JH, Williams SV, Krakauer H, Schwartz JS. Hospital and patient characteristics
associated with death after surgery. A study of adverse occurrence and failure to rescue.
Med Care. 1992;30:615-629.
14. Runciman WB, Sellen A, Webb RK, Williamson JA, Currie M, Morgan C, et al. The
Australian Incident Monitoring Study. Errors, incidents and accidents in anaesthetic
practice. Anaesth Intensive Care. 1993;21:506-519.
15. Donchin Y, Gopher D, Olin M, Badihi Y, Biesky M, Sprung CL, et al. A look into the
nature and causes of human errors in the intensive care unit. Critical Care Medicine.
1995;23:294-300.
16. Bates DW, Cullen DJ, Laird N, Petersen LA, Small SD, Servi D, et al. Incidence of adverse
drug events and potential adverse drug events. Implications for prevention. ADE
Prevention Study Group. JAMA. 1995;274:29-34.
17. Dean BS, Allan EL, Barber ND, Barker KN. Comparison of medication errors in an
American and a British hospital. Am J Health Syst Pharm. 1995;52:2543-2549.
18. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby L, Hamilton JD. The
Quality in Australian Health Care Study. Med J Aust. 1995;163:458-471.
19. Andrews LB, Stocking C, Krizek T, Gottlieb L, Krizek C, Vargish T, et al. An alternative
strategy for studying adverse events in medical care. Lancet. 1997;349:309-313.
20. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in
hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA.
1997;277:301-306.
21. Lesar TS, Lomaestro BM, Pohl H. Medication-prescribing errors in a teaching hospital. A
9-year experience. Arch Intern Med. 1997;157:1569-1576.
22. Thomas EJ, Studdert DM, Burstin HR, Orav EJ, Zeena T, Williams EJ, et al. Incidence and
types of adverse events and negligent care in Utah and Colorado. Med Care. 2000;38:261-
271.
22

23. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, et al. Incidence
of adverse events and negligence in hospitalized patients. Results of the Harvard Medical
Practice Study I. N Engl J Med. 1991;324:370-376.
24. Gandhi TK, Burstin HR, Cook EF, Puopolo AL, Haas JS, Brennan TA, et al. Drug
complications in outpatients. J Gen Intern Med. 2000;15:149-154.
25. Barach P, Small SD. Reporting and preventing medical mishaps: lessons from non-medical
near miss reporting systems. BMJ. 2000;320:759-763.
26. Leape LL. Reporting of medical errors: time for a reality check. Qual Health Care.
2000;9:144-145.
27. Dahl FC, Davis NM. A survey of hospital policies on verbal orders. Hosp Pharm.
1990;25:443-447.
28. West DW, Levine S, Magram G, MacCorkle AH, Thomas P, Upp K. Pediatric medication
order error rates related to the mode of order transmission. Arch Pediatr Adolesc Med.
1994;148:1322-1326.
29. Preliminary report: effect of encainide and flecainide on mortality in a randomized trial of
arrhythmia suppression after myocardial infarction. The Cardiac Arrhythmia Suppression
Trial (CAST) Investigators. N Engl J Med. 1989;321:406-412.
30. Hulley S, Grady D, Bush T, Furberg C, Herrington D, Riggs B, et al. Randomized trial of
estrogen plus progestin for secondary prevention of coronary heart disease in
postmenopausal women. Heart and Estrogen/progestin Replacement Study (HERS)
Research Group. JAMA. 1998;280:605-613.
31. Mennemeyer ST, Morrisey MA, Howard LZ. Death and reputation: how consumers acted
upon HCFA mortality information. Inquiry. 1997;34:117-128.
23

Chapter 2. Drawing on Safety Practices from Outside Health Care
The medical profession’s previous inattention to medical error, along with other
publicized deficiencies (such as a notable lag in adopting sophisticated information technologies)
have invited unfavorable comparisons between health care and other complex industries. 1-5 The
first of the two recent Institute of Medicine (IOM) reports on the quality of health care in
America, To Err is Human: Building a Safer Health System, 3 states that “health care is a decade
or more behind other high-risk industries in its attention to ensuring basic safety.” Consequently,
one of the goals of this project was to search these other industries for evidence-based safety
strategies that might be applied to health care.
The relatively short timeline of this project necessitated a focused approach to the search
for potentially applicable patient safety practices from non-health care writings. Fortunately,
many relevant practices have received at least some analysis or empirical study in the health care
literature. As a practical solution we present original articles from outside health care as
foundational and background material, rather than as a primary source of evidence. Specific
topics and practices reviewed in Making Health Care Safer that clearly derive from fields
outside health care include:
• Incident reporting (Chapter 4)
• Root cause analysis (Chapter 5)
• Computerized physician order entry and decision support as a means of
reducing medication errors (Chapter 6)
• Automated medication dispensing systems (Chapter 11)
• Bar coding technology to avoid misidentification errors (Subchapter 43.1)
• Aviation-style preoperative checklists for anesthesia equipment (Chapter 23
and Subchapter 41.3)
• Promoting a “culture of safety” (Chapter 40)
• Crew resource management, a model for teamwork training and crisis
response modeled after training approaches in aviation (Chapter 44)
• Simulators (of patients or clinical scenarios) as a training tool (Chapter 45)
• Human factors theory in the design of medical devices and alarms (Chapter
41)
Many readers may still wonder at the relative paucity of safety practices drawn from nonhealth
care sources. While the headline-grabbing assessments of medicine’s safety have been
criticized by researchers and likely overstate the hazard to patients, 6-8 it is undeniable that some
industries, most notably commercial aviation, have safety records far superior to that of health
care. One issue we faced in compiling this evidence-based review was the extent to which
specific practices could be identified as playing a direct and measurable role in this achievement.
Interestingly, the same issue—ascertaining a causative variable—arose in reviewing the
literature on anesthesia, likely the one field of medicine with a safety record that rivals aviation’s
(see also Chapter 56).
25

As outlined in Chapter 24, significant complications attributable to anesthesia have
decreased 9-12 to the point that major morbidity and mortality are now too rare to serve as
practical endpoints for measuring the quality of anesthesia care, even in large multicenter
studies. 13,14 In attempting to account for this decrease, however, it is very difficult to find
evidence supporting a causative role for even the most plausible candidates, such as widely
utilized intraoperative monitoring standards. 15 In other words, while the field of anesthesia has
clearly made tremendous strides in improving patient safety over the past 50 years, it is hard to
discern a particular, isolated practice that accounts for the clear and dramatic secular change in
its safety. While at one level, a pragmatist might argue, “who cares, as long as it’s safe,” trying
to adopt the lessons of anesthesia (or for that matter aviation) to the rest of health care is made
more challenging by tenuous causality.
Some might argue that, rather than pinpointing specific practices to embrace from other
industries, health care institutions should emulate organizational models that promote safety in
complex, high-risk industries that manage to operate with high reliability. 16 Analysis of detailed
and interesting case studies 17-22 have fueled a school of thought known as high reliability theory,
whose proponents suggest a number of organizational features that likely reduce the risk of
“organizational accidents” and other hazards. A cogently argued alternative position, often called
normal accident theory, questions not only these prescriptions for organizational change, but
fundamentally challenges the idea of high reliability in certain kinds of complex, “tightly
coupled” organizations. 23,24 These competing schools of thought offer interesting and valuable
insights into the ways that organizational strategies foster safety, while cautioning about the
ever-present threat of new sources of error that come with increasingly complex human and
technical organizations. Unfortunately, this rich literature does not permit ready synthesis within
the framework of evidence-based medicine, even using the less stringent standards we adopted in
evaluating non-medical literature (see Chapters 1 and 3).
Even the more engineering-oriented of the disciplines with potential relevance to patient
safety yielded a surprising lack of empirical evaluation of safety practices. For instance,
numerous techniques for “human error identification” and “error mode prediction” purport to
anticipate important errors and develop preventive measures prospectively. 25-27 Their basic
approach consists of breaking down the task of interest into component processes, and then
assigning a measure of the likelihood of failure to each process. Many of the techniques
mentioned in the literature have received little detailed description 25,26 and few have received
any formal validation (eg, by comparing predicted failures modes with observed errors). 28,29
Even setting aside demands for validation, the impact of applying these techniques has not been
assessed. Total quality management and continuous quality improvement techniques were
championed as important tools for change in health care based on their presumed success in
other industries, but evaluations of their impact on health care have revealed little evidence of
success. 30-33
In the end, we are left with our feet firmly planted in the middle of competing paradigms.
One argues that an evidence-based, scientific approach has served health care well and should
not be relaxed simply because a popular practice from a “safer” industry sounds attractive. The
other counters that medicine’s slavish devotion to the scientific and epidemiologic method has
placed us in a patient safety straightjacket, unable to consider the value of practices developed in
other fields because of our myopic traditions and “reality.”
We see the merits in both arguments. Health care clearly has much to learn from other
industries. Just as physicians must learn the “basic sciences” of immunology and molecular
biology, providers and leaders interested in making health care safer must learn the “basic
26

sciences” of organizational theory and human factors engineering. Moreover, the “cases”
presented on rounds should, in addition to classical clinical descriptions, also include the tragedy
of the Challenger and the successes of Motorola. On the other hand, an unquestioning embrace
of dozens of promising practices from other fields is likely to be wasteful, distracting, and
potentially dangerous. We are drawn to a dictum from the Cold War era—“Trust, but verify.”
References
1. Leape LL. Error in medicine. JAMA. 1994;272:1851-1857.
2. Chassin MR. Is health care ready for Six Sigma quality? Milbank Q. 1998;76:565-591,510.
3. Kohn L, Corrigan J, Donaldson M, editors. To Err Is Human: Building a Safer Health
System. Washington, DC: Committee on Quality of Health Care in America, Institute of
Medicine. National Academy Press; 2000.
4. Barach P, Small SD. Reporting and preventing medical mishaps: lessons from non-medical
near miss reporting systems. BMJ. 2000;320:759-763.
5. Helmreich RL. On error management: lessons from aviation. BMJ. 2000;320:781-785.
6. Brennan TA. The Institute of Medicine report on medical errors–could it do harm? N Engl
J Med. 2000;342:1123-1125.
7. McDonald CJ, Weiner M, Hui SL. Deaths due to medical errors are exaggerated in Institute
of Medicine report. JAMA. 2000;284:93-95.
8. Sox HCJ, Woloshin S. How many deaths are due to medical error? Getting the number
right. Eff Clin Pract. 2000;3:277-283.
9. Phillips OC, Capizzi LS. Anesthesia mortality. Clin Anesth. 1974;10:220-244.
10. Deaths during general anesthesia: technology-related, due to human error, or unavoidable?
An ECRI technology assessment. J Health Care Technol. 1985;1:155-175.
11. Keenan RL, Boyan CP. Cardiac arrest due to anesthesia. A study of incidence and causes.
JAMA. 1985;253:2373-2377.
12. Eichhorn JH. Prevention of intraoperative anesthesia accidents and related severe injury
through safety monitoring. Anesthesiology. 1989;70:572-577.
13. Cohen MM, Duncan PG, Pope WD, Biehl D, Tweed WA, MacWilliam L, et al. The
Canadian four-centre study of anaesthetic outcomes: II. Can outcomes be used to assess the
quality of anaesthesia care? Can J Anaesth. 1992;39:430-439.
14. Moller JT, Johannessen NW, Espersen K, Ravlo O, Pedersen BD, Jensen PF, et al.
Randomized evaluation of pulse oximetry in 20,802 patients: II. Perioperative events and
postoperative complications. Anesthesiology. 1993;78:445-453.
15. Orkin FK. Practice standards: the Midas touch or the emperor's new clothes?
Anesthesiology. 1989;70:567-571.
16. Reason J. Human error: models and management. BMJ. 2000;320:768-770.
17. Weick KE. Organizational culture as a source of high reliability. California management
review. 1987;29:112-127.
18. Roberts KH. Some characteristics of high reliability organizations. Berkeley, CA:
Produced and distributed by Center for Research in Management, University of California,
Berkeley Business School; 1988.
19. LaPorte TR. The United States air traffic control system : increasing reliability in the midst
of rapid growth. In: Mayntz R, Hughes TP, editors. The Development of Large technical
Systems. Boulder, CO: Westview Press; 1988.
20. Roberts KH. Managing High Reliability Organizations. California Management Review.
1990;32:101-113.
27

21. Roberts KH, Libuser C. From Bhopal to banking: Organizational design can mitigate risk.
Organizational Dynamics. 1993;21:15-26.
22. LaPorte TR, Consolini P. Theoretical and operational challenges of "high-reliability
organizations": Air-traffic control and aircraft carriers. International Journal of Public
Administration. 1998;21:847-852.
23. Perrow C. Normal accidents : Living with High-Risk Technologies. With a New Afterword
and a Postscript on the Y2K Problem. Princeton, NJ: Princeton University Press; 1999.
24. Sagan SD. The Limits of Safety: Organizations, Accidents and Nuclear Weapons.
Princeton, NJ: Princeton University Press; 1993.
25. Kirwan B. Human error identification techniques for risk assessment of high risk systems–
Part 1: Review and evaluation of techniques. Appl Ergon. 1998;29:157-177.
26. Kirwan B. Human error identification techniques for risk assessment of high risk systems–
Part 2: towards a framework approach. Appl Ergon. 1998;29:299-318.
27. Hollnagel E, Kaarstad M, Lee HC. Error mode prediction. Ergonomics. 1999;42:1457-
1471.
28. Kirwan B, Kennedy R, Taylor-Adams S, Lambert B. The validation of three human
reliability quantification techniques–THERP, HEART and JHEDI: Part II–Results of
validation exercise. Appl Ergon. 1997;28:17-25.
29. Stanton NA, Stevenage SV. Learning to predict human error: issues of acceptability,
reliability and validity. Ergonomics. 1998;41:1737-1756.
30. Blumenthal D, Kilo CM. A report card on continuous quality improvement. Milbank Q.
1998;76:625-48,511.
31. Gerowitz MB. Do TQM interventions change management culture? Findings and
implications. Qual Manag Health Care. 1998;6:1-11.
32. Shortell SM, Bennett CL, Byck GR. Assessing the impact of continuous quality
improvement on clinical practice: what it will take to accelerate progress. Milbank Q.
1998;76:593-624,510.
33. Shortell SM, Jones RH, Rademaker AW, Gillies RR, Dranove DS, Hughes EF, et al.
Assessing the impact of total quality management and organizational culture on multiple
outcomes of care for coronary artery bypass graft surgery patients. Med Care.
2000;38:207-217.
28

Chapter 3. Evidence-based Review Methodology
Definition and Scope
For this project the UCSF-Stanford Evidence-based Practice Center (EPC) defined a
patient safety practice as “a type of process or structure whose application reduces the
probability of adverse events resulting from exposure to the health care system across a range of
conditions or procedures.” Examples of practices that meet this definition include computerized
physician order entry (Chapter 6), thromboembolism prophylaxis in hospitalized patients
(Chapter 31), strategies to reduce falls among hospitalized elders (Chapter 26), and novel
education strategies such as the application of “crew resource management” to train operating
room staff (Chapter 44). By contrast, practices that are disease-specific and/or directed at the
underlying disease and its complications (eg, use of aspirin or beta-blockers to treat acute
myocardial infarction) rather than complications of medical care are not included as “patient
safety practices.” Further discussion of these issues can be found in Chapter 1.
In some cases, the distinction between patient safety and more general quality
improvement strategies is difficult to discern. Quality improvement practices may also qualify as
patient safety practices when the current level of quality is considered “unsafe,” but standards to
measure safety are often variable, difficult to quantify, and change over time. For example, what
constitutes an “adequate” or “safe” level of accuracy in electrocardiogram or radiograph
interpretation? Practices to improve performance to at least the “adequate” threshold may
reasonably be considered safety practices because they decrease the number of diagnostic errors
of omission. On the other hand, we considered practices whose main intent is to improve
performance above this threshold to be quality improvement practices. An example of the latter
might be the use of computer algorithms to improve the sensitivity of screening mammography. 1
We generally included practices that involved acute hospital care or care at the interface
between inpatient and outpatient settings. This focus reflects the fact that the majority of the
safety literature relates to acute care and the belief that systems changes may be more effectively
achieved in the more controlled environment of the hospital. However, practices that might be
applicable in settings in addition to the hospital were not excluded from consideration. For
example, the Report includes practices for preventing decubitus ulcers (Chapter 27) that could be
applied in nursing homes as well as hospitals.
The EPC team received input regarding the scope of the Report from the Agency for
Healthcare Research and Quality (AHRQ), which commissioned the report. In addition, the EPC
team participated in the public meeting of the National Quality Forum (NQF) Safe Practices
Committee on January 26, 2001. The NQF was formed in 1999 by consumer, purchaser,
provider, health plan, and health service research organizations to create a national strategy for
quality improvement. Members of the Safe Practices Committee collaborated with the EPC team
to develop the scope of work that would eventually become this Report.
Organization by Domains
To facilitate identification and evaluation of potential patient safety practices, we divided
the content for the project into different domains. Some cover “content areas” (eg, adverse drug
events, nosocomial infections, and complications of surgery). Other domains involve
identification of practices within broad (primarily “non-medical”) disciplines likely to contain
promising approaches to improving patient safety (eg, information technology, human factors
research, organizational theory). The domains were derived from a general reading of the
29

literature and were meant to be as inclusive as possible. The list underwent review for
completeness by patient safety experts, clinician-researchers, AHRQ, and the NQF Safe
Practices Committee. For each domain we selected a team of author/collaborators with expertise
in the relevant subject matter and/or familiarity with the techniques of evidence-based review
and technology appraisal. The authors, all of whom are affiliated with major academic centers
around the United States, are listed on page 9-13.
Identification of Safety Practices for Evaluation
Search Strategy
The UCSF-Stanford Evidence-based Practice Center Coordinating Team (“The Editors”)
provided general instructions (Table 3.1) to the teams of authors regarding search strategies for
identifying safety practices for evaluation. As necessary, the Editors provided additional
guidance and supplementary searches of the literature.
Table 3.1. Search strategy recommended by coordinating team to
participating reviewers
a) Electronic bibliographic databases. All searches must include systematic
searches of MEDLINE and the Cochrane Library. For many topics it will
be necessary to include other databases such as the Cumulative Index to
Nursing & Allied Health (CINAHL), PsycLit (PsycINFO), the Institute for
Scientific Information’s Science Citation Index Expanded, Social Sciences
Citation Index, Arts & Humanities Citation Index, INSPEC (physics,
electronics and computing), and ABI/INFORM (business, management,
finance, and economics).
b) Hand-searches of bibliographies of retrieved articles and tables of
contents of key journals.
c) Grey literature. For many topics it will necessary to review the “grey
literature,” such as conference proceedings, institutional reports, doctoral
theses, and manufacturers’ reports.
d) Consultation with experts or workers in the field.
To meet the early dissemination date mandated by AHRQ, the Editors did not require
authors to search for non-English language articles or to use EMBASE. These were not
specifically excluded, however, and authoring teams could include non-English language articles
that addressed important aspects of a topic if they had translation services at their disposal. The
Editors did not make recommendations on limiting database searches based on publication date.
For this project it was particularly important to identify systematic reviews related to patient
safety topics. Published strategies for retrieving systematic reviews have used proprietary
MEDLINE interfaces (eg, OVID, SilverPlatter) that are not uniformly available. Moreover, the
performance characteristics of these search strategies is unknown. 2,3 Therefore, these strategies
were not explicitly recommended. The Editors provided authors with a search algorithm
(available upon request) that uses PubMed, the freely available search interface from the
National Library of Medicine, designed to retrieve systematic reviews with high sensitivity
without overwhelming users with “false positive” hits. 4
30

The Editors also performed independent searches of bibliographic databases and grey
literature for selected topics. 5 Concurrently, the EPC collaborated with NQF to solicit
information about evidence-based practices from NQF members, and consulted with the
project’s Advisory Panel (page 31), whose members provided additional literature to review.
Inclusion and Exclusion Criteria
The EPC established criteria for selecting which of the identified safety practices
warranted evaluation. The criteria address the applicability of the practice across a range of
conditions or procedures and the available evidence of the practices’ efficacy or effectiveness.
Table 3.2. Inclusion/Exclusion criteria for practices
Inclusion Criteria
1. The practice can be applied in the hospital setting or at the interface
between inpatient and outpatient settings AND can be applied to a broad
range of health care conditions or procedures.
2. Evidence for the safety practice includes at least one study with a Level 3
or higher study design AND a Level 2 outcome measure. For practices not
specifically related to diagnostic or therapeutic interventions, a Level 3
outcome measure is adequate. (See Table 3.3 for definition of “Levels”).
Exclusion Criterion
1. No study of the practice meets the methodologic criteria above.
Practices that have only been studied outside the hospital setting or in patients with
specific conditions or undergoing specific procedures were included if the authors and Editors
agreed that the practices could reasonably be applied in the hospital setting and across a range of
conditions or procedures. To increase the number of potentially promising safety practices
adapted from outside the field of medicine, we included evidence from studies that used less
rigorous measures of patient safety as long as the practices did not specifically relate to
diagnostic or therapeutic interventions. These criteria facilitated the inclusion of areas such as
teamwork training (Chapter 44) and methods to improve information transfer (Chapter 42).
Each practice’s level of evidence for efficacy or effectiveness was assessed in terms of
study design (Table 3.3) and study outcomes (Table 3.4). The Editors created the following
hierarchies by modifying existing frameworks for evaluating evidence 6-8 and incorporating
recommendations from numerous other sources relevant to evidence synthesis. 9-25
31

Table 3.3. Hierarchy of study designs*
Level 1. Randomized controlled trials – includes quasi-randomized processes
such as alternate allocation
Level 2. Non-randomized controlled trial – a prospective (pre-planned) study,
with predetermined eligibility criteria and outcome measures.
Level 3. Observational studies with controls – includes retrospective,
interrupted time series (a change in trend attributable to the
intervention), case-control studies, cohort studies with controls, and
health services research that includes adjustment for likely
confounding variables
Level 4. Observational studies without controls (eg, cohort studies without
controls and case series)
* Systematic reviews and meta-analyses were assigned to the highest level
study design included in the review, followed by an “A” (eg, a systematic
review that included at least one randomized controlled trial was designated
“Level 1A”)
Table 3.4. Hierarchy of outcome measures
Level 1. Clinical outcomes - morbidity, mortality, adverse events
Level 2. Surrogate outcomes - observed errors, intermediate outcomes (eg,
laboratory results) with well-established connections to the clinical
outcomes of interest (usually adverse events).
Level 3. Other measurable variables with an indirect or unestablished
connection to the target safety outcome (eg, pre-test/post-test after
an educational intervention, operator self-reports in different
experimental situations)
Level 4. No outcomes relevant to decreasing medical errors and/or adverse
events (eg, study with patient satisfaction as only measured
outcome; article describes an approach to detecting errors but reports
no measured outcomes)
Implicit in this hierarchy of outcome measures is that surrogate or intermediate outcomes
(Level 2) have an established relationship to the clinical outcomes (Level 1) of interest. 26
Outcomes that are relevant to patient safety but have not been associated with morbidity or
mortality were classified as Level 3.
Exceptions to EPC Criteria
Some safety practices did not meet the EPC inclusion criteria because of the paucity of
evidence regarding efficacy or effectiveness, but were included in the Report because of their
face validity (ie, an informed reader might reasonably expect them to be evaluated; see also
Chapter 1). The reviews of these practices clearly identify the quality of evidence culled from
medical and non-medical fields.
32

Evaluation of Safety Practices
For each practice, authors were instructed to research the literature for information on:
• prevalence of the problem targeted by the practice
• severity of the problem targeted by the practice
• the current utilization of the practice
• evidence on efficacy and/or effectiveness of the practice
• the practice’s potential for harm
• data on cost if available
• implementation issues
These elements were incorporated into a template in an effort to create as much
uniformity across chapters as possible, especially given the widely disparate subject matter and
quality of evidence. Since the amount of material for each practice was expected to, and did,
vary substantially, the Editors provided general guidance on what was expected for each
element, with particular detail devoted to the protocol for searching and reporting evidence
related to efficacy and/or effectiveness of the practice.
The protocol outlined the search, the threshold for study inclusion, the elements to
abstract from studies, and guidance on reporting information from each study. Authors were
asked to review articles from their search to identify practices, and retain those with the better
study designs. More focused searches were performed depending on the topic. The threshold for
study inclusion related directly to study design. Authors were asked to use their judgment in
deciding whether the evidence was sufficient at a given level of study design or whether the
evidence from the next level needed to be reviewed. At a minimum, the Editors suggested that
there be at least 2 studies of adequate quality to justify excluding discussion of studies of lower
level designs. Thus inclusion of 2 adequate clinical trials (Level 1 design) were necessary in
order to exclude available evidence from prospective, non-randomized trials (Level 2) on the
same topic.
The Editors provided instructions for abstracting each article that met the inclusion
criteria based on study design. For each study, a required set of 10 abstraction elements (Table
3.5) was specified. Authors received a detailed explanation of each required abstraction element,
as well as complete abstraction examples for the 3 types of study design (Levels 1A, 1, and 3;
Level 2 was not included since the information collected was same as Level 1). Research teams
were encouraged to abstract any additional elements relevant to the specific subject area.
33

Table 3.5. Ten Required Abstraction Elements
1. Bibliographic information according to AMA Manual of Style: title, authors, date
of publication, source
2. Level of study design (eg, Level 1-3 for studies providing information for
effectiveness; Level 4 if needed for relevant additional information) with
descriptive material as follows:
Descriptors
For Level 1 Systematic Reviews Only
(a) Identifiable description of methods indicating sources and methods of
searching for articles.
(b) Stated inclusion and exclusion criteria for articles: yes/no.
(c) Scope of literature included in study.
For Level 1 or 2 Study Designs (Not Systematic Reviews)
(a) Blinding: blinded, blinded (unclear), or unblinded
(b) Describe comparability of groups at baseline – ie, was distribution of potential
confounders at baseline equal? If no, which confounders were not equal?
(c) Loss to follow-up overall: percent of total study population lost to follow-up.
For Level 3 Study Design (Not Systematic Reviews)
(a) Description of study design (eg, case-control, interrupted time series).
(b) Describe comparability of groups at baseline – ie, was distribution of potential
confounders at baseline equal? If no, which confounders were not equal?
(c) Analysis includes adjustment for potential confounders: yes/no. If yes,
adjusted for what confounders?
3. Description of intervention (as specific as possible)
4. Description of study population(s) and setting(s)
5. Level of relevant outcome measure(s) (eg, Levels 1-4)
6. Description of relevant outcome measure(s)
7. Main Results: effect size with confidence intervals
8. Information on unintended adverse (or beneficial) effects of practice
9. Information on cost of practice
10. Information on implementation of practice (information that might be of use in
whether to and/or how to implement the practice – eg, known barriers to
implementation)
We present the salient elements of each included study (eg, study design,
population/setting, intervention details, results) in text or tabular form. In addition, we asked
authors to highlight weaknesses and biases of studies where the interpretation of the results
might be substantially affected. Authors were not asked to formally synthesize or combine (eg,
perform a meta-analysis) the evidence across studies for the Report.
34

Review Process
Authors submitted work to the Editors in 2 phases. In the first phase (“Identification of
Safety Practices for Evaluation”), which was submitted approximately 6 weeks after authors
were commissioned, authoring teams provided their search strategies, citations, and a
preliminary list of patient safety practices to be reviewed. In the subsequent phase (“Evaluation
of Safety Practices”), due approximately 12 weeks after commissioning, authors first submitted a
draft chapter for each topic, completed abstraction forms, and—after iterative reviews and
revisions—a final chapter.
Identification of Safety Practices for Evaluation
The Editors and the Advisory Panel reviewed the list of domains and practices to identify
gaps in coverage. In addition, the Editors reviewed final author-submitted lists of excluded
practices along with justifications for exclusion (eg, insufficient research design, insufficient
outcomes, practice is unique to a single disease process). When there were differences in opinion
as to whether a practice actually met the inclusion or exclusion criteria, the Editors made a final
disposition after consulting with the author(s). The final practice list, in the form of a Table of
Contents for the Report, was circulated to AHRQ and the NQF Safe Practices Committee for
comment.
Evaluation of Safety Practices
Chapters were reviewed by the editorial team (The EPC Coordinating Team Editors and
our Managing Editor) and queries were relayed to the authors, often requesting further
refinement of the analysis or expansion of the results and conclusions. After all chapters were
completed, the entire Report was edited to eliminate redundant material and ensure that the focus
remained on the evidence regarding safety practices. Near the end of the review process,
chapters were distributed to the Advisory Panel for comments, many of which were
incorporated. Once the content was finalized, the Editors analyzed and ranked the practices using
a methodology described in Chapter 56. The results of these summaries and rankings are
presented in Part V of the Report.
35

References
1. Thurfjell E, Thurfjell MG, Egge E, Bjurstam N. Sensitivity and specificity of computerassisted
breast cancer detection in mammography screening. Acta Radiol. 1998;39:384-
388.
2. Hunt DL, McKibbon KA. Locating and appraising systematic reviews Ann Intern Med.
1997;126:532-538.
3. Glanville J, Lefebvre C. Identifying systematic reviews: key resources. ACP J Club.
2000;132:A11-A12.
4. Shojania KG, Bero L. Taking advantage of the explosion of systematic reviews: an
efficient MEDLINE search strategy. Eff Clin Pract. 2001;4: in press.
5. McAuley L, Pham B, Tugwell P, Moher D. Does the inclusion of grey literature influence
estimates of intervention effectiveness reported in meta-analyses? Lancet. 2000;356:1228-
1231.
6. Harris RP, Helfand M, Woolf SH, Lohr KN, Mulrow CD, Teutsch SM, et al. Current
methods of the U.S. Preventive Services Task Force. A review of the process. Am J Prev
Med. 2001;20:21-35.
7. Guyatt G, Schunemann H, Cook D, Jaeschke R, Pauker S, Bucher H. Grades of
recommendation for antithrombotic agents. Chest.2001;119:3S-7S.
8. Muir Gray JA, Haynes RB, Sackett DL, Cook DJ, Guyatt GH. Transferring evidence from
research into practice: 3. Developing evidence-based clinical policy. ACP J Club.
1997;126:A14-A16.
9. Guyatt GH, Tugwell PX, Feeny DH, Drummond MF, Haynes RB. The role of before-after
studies of therapeutic impact in the evaluation of diagnostic technologies. J Chronic Dis.
1986;39:295-304.
10. Davis CE. Generalizing from clinical trials. Control Clin Trials. 1994;15:11-4.
11. Bailey KR. Generalizing the results of randomized clinical trials. Control Clin Trials.
1994;15:15-23.
12. Oxman AD, Cook DJ, Guyatt GH. Users' guides to the medical literature. VI. How to use
an overview. Evidence-Based Medicine Working Group. JAMA. 1994;272:1367-1371.
13. Levine M, Walter S, Lee H, Haines T, Holbrook A, Moyer V. Users' guides to the medical
literature. IV. How to use an article about harm. Evidence-Based Medicine Working
Group. JAMA. 1994;271:1615-1619.
14. Guyatt GH, Sackett DL, Sinclair JC, Hayward R, Cook DJ, Cook RJ. Users' guides to the
medical literature. IX. A method for grading health care recommendations. Evidence-
Based Medicine Working Group. JAMA. 1995;274:1800-1804.
15. Cook DJ, Guyatt GH, Laupacis A, Sackett DL, Goldberg RJ. Clinical recommendations
using levels of evidence for antithrombotic agents. Chest. 1995;108:227S-230S.
16. Naylor CD, Guyatt GH. Users' guides to the medical literature. X. How to use an article
reporting variations in the outcomes of health services. The Evidence-Based Medicine
Working Group. JAMA. 1996;275:554-558.
17. Moher D, Jadad AR, Tugwell P. Assessing the quality of randomized controlled trials.
Current issues and future directions. Int J Technol Assess Health Care. 1996;12:195-208.
18. Mulrow C, Langhorne P, Grimshaw J. Integrating heterogeneous pieces of evidence in
systematic reviews. Ann Intern Med. 1997;127:989-995.
19. Jadad AR, Cook DJ, Browman GP. A guide to interpreting discordant systematic reviews.
CMAJ. 1997;156:1411-1416.
36

20. Juni P, Witschi A, Bloch R, Egger M. The hazards of scoring the quality of clinical trials
for meta-analysis. JAMA. 1999;282:1054-1060.
21. McKee M, Britton A, Black N, McPherson K, Sanderson C, Bain C. Methods in health
services research. Interpreting the evidence: choosing between randomised and nonrandomised
studies. BMJ. 1999;319:312-315.
22. Bucher HC, Guyatt GH, Cook DJ, Holbrook A, McAlister FA. Users' guides to the medical
literature: XIX. Applying clinical trial results. A. How to use an article measuring the
effect of an intervention on surrogate end points. Evidence-Based Medicine Working
Group. JAMA. 1999;282:771-778.
23. Concato J, Shah N, Horwitz RI. Randomized, controlled trials, observational studies, and
the hierarchy of research designs. N Engl J Med. 2000;342:1887-1892.
24. Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled
trials. N Engl J Med. 2000;342:1878-1886.
25. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis
of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of
Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283:2008-2012.
26. Gotzsche PC, Liberati A, Torri V, Rossetti L. Beware of surrogate outcome measures. Int J
Technol Assess Health Care. 1996;12:238-246.
37

PART II. REPORTING AND RESPONDING TO PATIENT SAFETY PROBLEMS
Chapter 4. Incident Reporting
Chapter 5. Root Cause Analysis
39

Chapter 4. Incident Reporting
Heidi Wald, MD
University of Pennsylvania School of Medicine
Kaveh G. Shojania, MD
University of California, San Francisco School of Medicine
Background
Errors in medical care are discovered through a variety of mechanisms. Historically,
medical errors were revealed retrospectively through morbidity and mortality committees and
malpractice claims data. Prominent studies of medical error have used retrospective chart review
to quantify adverse event rates. 1,2 While collection of data in this manner yields important
epidemiologic information, it is costly and provides little insight into potential error reduction
strategies. Moreover, chart review only detects documented adverse events and often does not
capture information regarding their causes. Important errors that produce no injury may go
completely undetected by this method. 3-6
Computerized surveillance may also play a role in uncovering certain types of errors. For
instance, medication errors may be discovered through a search for naloxone orders for
hospitalized patients, as they presumably reflect the need to reverse overdose of prescribed
narcotics. 7,8 Several studies have demonstrated success with computerized identification of
adverse drug events. 9-11
Complex, high-risk industries outside of health care, including aviation, nuclear power,
petrochemical processing, steel production, and military operations, have successfully developed
incident reporting systems for serious accidents and important “near misses.” 6 Incident reporting
systems cannot provide accurate epidemiologic data, as the reported incidents likely
underestimate the numerator, and the denominator (all opportunities for incidents) remains
unknown.
Given the limited availability of sophisticated clinical computer systems and the
tremendous resources required to conduct comprehensive chart reviews, incident reporting
systems remain an important and relatively inexpensive means of capturing data on errors and
adverse events in medicine. Few rigorous studies have analyzed the benefits of incident
reporting. This chapter reviews only the literature evaluating the various systems and techniques
for collecting error data in this manner, rather than the benefit of the practice itself. This decision
reflects our acknowledgment that incident reporting has clearly played a beneficial role in other
high-risk industries. 6 The decision also stems from our recognition that a measurable impact of
incident reporting on clinical outcomes is unlikely because there is no standard practice by which
institutions handle these reports.
Practice Description
Flanagan first described the critical incident technique in 1954 to examine military
aircraft training accidents. 12 Critical incident reporting involves the identification of preventable
incidents (ie, occurrences that could have led, or did lead, to an undesirable outcome 13 ) reported
by personnel directly involved in the process in question at the time of discovery of the event.
The goal of critical incident monitoring is not to gather epidemiologic data per se, but rather to
gather qualitative data. Nonetheless, if a pattern of errors seems to emerge, prospective studies
can be undertaken to test epidemiologic hypotheses. 14
41

Incident reports may target events in any or all of 3 basic categories: adverse events, “no
harm events,” and “near misses.” For example, anaphylaxis to penicillin clearly represents an
adverse event. Intercepting the medication order prior to administration would constitute a near
miss. By contrast, if a patient with a documented history of anaphylaxis to penicillin received a
penicillin-like antibiotic (eg, a cephalosporin) but happened not to experience an allergic
reaction, it would constitute a no harm event, not a near miss. In other words, when an error does
not result in an adverse event for a patient, because the error was “caught,” it is a near miss; if
the absence of injury is owed to chance it is a no harm event. Broadening the targets of incident
reporting to include no harm events and near misses offers several advantages. These events
occur 3 to 300 times more often than adverse events, 5,6 they are less likely to provoke guilt or
other psychological barriers to reporting, 6 and they involve little medico-legal risk. 14 In addition,
hindsight bias 15 is less likely to affect investigations of no harm events and near misses. 6,14
Barach and Small describe the characteristics of incident reporting systems in nonmedical
industries. 6 Established systems share the following characteristics:
• they focus on near misses
• they provide incentives for voluntary reporting;
• they ensure confidentiality; and
• they emphasize systems approaches to error analysis.
The majority of these systems were mandated by Federal regulation, and provide for
voluntary reporting. All of the systems encourage narrative description of the event. Reporting is
promoted by providing incentives including:
• immunity;
• confidentiality;
• outsourcing of report collation;
• rapid feedback to all involved and interested parties; and
• sustained leadership support. 6
Incident reporting in medicine takes many forms. Since 1975, the US Food and Drug
Administration (FDA) has mandated reporting of major blood transfusion reactions, focusing on
preventable deaths and serious injuries. 16 Although the critical incident technique found some
early applications in medicine, 17,18 its current use is largely attributable to Cooper’s introduction
of incident reporting to anesthesia in 1978, 19 conducting retrospective interviews with
anesthesiologists about preventable incidents or errors that occurred while patients were under
their care. Recently, near miss and adverse event reporting systems have proliferated in single
institution settings (such as in intensive care units (ICUs) 20,21 ), regional settings (such as the New
York State transfusion system 22 ), and for national surveillance (eg, the National Nosocomial
Infections Surveillance System administered by the Federal Centers for Disease Control and
Prevention.) 23
All of the above examples focus on types of events (transfusion events or nosocomial
infections) or areas of practice (ICUs). Incident reporting in hospitals cuts a wider swath,
capturing errors and departures from expected procedures or outcomes (Table 4.1). However,
because risk management departments tend to oversee incident reporting systems in some
capacity, these systems more often focus on incident outcomes, not categories. Few data describe
the operation of these institution-specific systems, but underreporting appears endemic. 24
42

In 1995, hospital-based surveillance was mandated by the Joint Commission on the
Accreditation of Healthcare Organizations (JCAHO) 26 because of a perception that incidents
resulting in harm were occurring frequently. 28 JCAHO employs the term sentinel event in lieu of
critical incident, and defines it as follows:
An unexpected occurrence involving death or serious physical or psychological
injury, or the risk thereof. Serious injury specifically includes loss of limb or
function. The phrase “or the risk thereof” includes any process variation for
which a recurrence would carry a significant chance of a serious adverse
outcome. 26
As one component of its Sentinel Event Policy, JCAHO created a Sentinel Event
Database. The JCAHO database accepts voluntary reports of sentinel events from member
institutions, patients and families, and the press. 26 The particulars of the reporting process are left
to the member health care organizations. JCAHO also mandates that accredited hospitals
perform root cause analysis (see Chapter 5) of important sentinel events. Data on sentinel events
are collated, analyzed, and shared through a website, 29 an online publication, 30 and its newsletter
Sentinel Event Perspectives. 31
Another example of a national incident reporting system is the Australian Incident
Monitoring Study (AIMS), under the auspices of the Australian Patient Safety Foundation. 13
Investigators created an anonymous and voluntary near miss and adverse event reporting system
for anesthetists in Australia. Ninety participating hospitals and practices named on-site
coordinators. The AIMS group developed a form that was distributed to participants. The form
contained instructions, definitions, space for narrative of the event, and structured sections to
record the anesthesia and procedure, demographics about the patient and anesthetist, and what,
when, why, where, and how the event occurred. The results of the first 2000 reports were
published together, following a special symposium. 32
The experiences of the JCAHO Sentinel Event Database and the Australian Incident
Monitoring Study are explored further below.
Prevalence and Severity of the Target Safety Problem
The true prevalence of events appropriate for incident reporting is impossible to estimate with
any accuracy, as it includes actual adverse events as well as near misses and no harm events. The
Aviation Safety Reporting System (ASRS), a national reporting system for near misses in the
airline industry,33,34 currently processes approximately 30,000 reports annually,35 exceeding
by many orders of magnitude the total number of airline accidents each year.34 The number of
reports submitted to a comparable system in health care would presumably number in the
millions if all adverse events, no harm events, and near misses were captured.
By contrast, over 6 years of operation, the JCAHO Sentinel Event Database has captured
only 1152 events, 62% of which occurred in general hospitals. Two-thirds of the events were
self-reported by institutions, with the remainder coming from patient complaints, media stories
and other sources. 29 These statistics are clearly affected by underreporting and consist primarily
of serious adverse events (76% of events reported resulted in patient deaths), not near misses. As
discussed in the chapter on wrong-site surgeries (Subchapter 43.2), comparing JCAHO reports
with data from the mandatory incident reporting system maintained by the New York State
Department of Health 36 suggests that the JCAHO statistics underestimate the true incidence of
target events by at least a factor of 20.
43

Opportunities for Impact
Most hospitals’ incident reporting systems fail to capture the majority of errors and near
misses. 24 Studies of medical services suggest that only 1.5% of all adverse events result in an
incident report 37 and only 6% of adverse drug events are identified through traditional incident
reporting or a telephone hotline. 24 The American College of Surgeons estimates that incident
reports generally capture only 5-30% of adverse events. 38 A study of a general surgery service
showed that only 20% of complications on a surgical service ever resulted in discussion at
Morbidity and Mortality rounds. 39 Given the endemic underreporting revealed in the literature,
modifications to the configuration and operation of the typical hospital reporting system could
yield higher capture rates of relevant clinical data.
Study Designs
We analyzed 5 studies that evaluated different methods of critical incident reporting.
Two studies prospectively investigated incident reporting compared with observational data
collection 24,39 and one utilized retrospective chart review. 37 Two additional studies looked at
enhanced incident reporting by active solicitation of physician input compared with background
hospital quality assurance (QA) measures. 40,41 In addition, we reviewed JCAHO’s report of its
Sentinel Event Database, and the Australian Incident Monitoring Study, both because of the
large sizes and the high profiles of the studies. 13,26 Additional reports of critical incident
reporting systems in the medical literature consist primarily of uncontrolled observational
trials 42-44 that are not reviewed in this chapter.
Study Outcomes
In general, published studies of incident reporting do not seek to establish the benefit of
incident reporting as a patient safety practice. Their principal goal is to determine if incident
reporting, as it is practiced, captures the relevant events. 40 In fact, no studies have established the
value of incident reporting on patient safety outcomes.
The large JCAHO and Australian databases provide data about reporting rates, and an
array of quantitative and qualitative information about the reported incidents, including the
identity of the reporter, time of report, severity and type of error. 13,26 Clearly these do not
represent clinical outcomes, but they may be reasonable surrogates for the organizational focus
on patient safety. For instance, increased incident reporting rates may not be indicative of an
unsafe organization, 45 but may reflect a shift in organizational culture to increased acceptance of
quality improvement and other organizational changes. 3,5
None of the studies reviewed captured outcomes such as morbidity or error rates. The
AIMS group published an entire symposium which reported the quantitative and qualitative data
regarding 2000 critical incidents in anesthesia. 13 However, only a small portion of these
incidents were prospectively evaluated. 14 None of the studies reviewed for this chapter
performed formal root cause analyses on reported errors (Chapter 5).
Evidence for Effectiveness of the Practice
As described above, 6 years of JCAHO sentinel event data have captured merely 1152
events, none of which include near misses. 29 Despite collecting what is likely to represent only a
fraction of the target events, JCAHO has compiled the events, reviewed the root cause analyses
and provided recommendations for procedures to improve patient safety for events ranging from
wrong-site surgeries to infusion pump-related adverse events (Table 4.2). This information may
44

prove to be particularly useful in the case of rare events such as wrong-site surgery, where
national collection of incidents can yield a more statistically useful sample size.
The first 2000 incident reports to AIMS from 90 member institutions were published in
1993. 13 In contrast to the JCAHO data, all events were self-reported by anesthetists and only 2%
of events reported resulted in patient deaths. A full 44% of events had negligible effect on
patient outcome. Ninety percent of reports had identified systems failures, and 79% had
identified human failures. The AIMS data were similar to those of Cooper 19 in terms of percent
of incidents with reported human failures, timing of events with regard to phase of anesthesia,
and type of events (breathing circuit misconnections were between 2% and 3% in both
studies). 13,19 The AIMS data are also similar to American “closed-claims” data in terms of
pattern, nature and proportion of the total number of reports for several types of adverse events, 13
which lends further credibility to the reports.
The AIMS data, although also likely to be affected by underreporting because of its
voluntary nature, clearly captures a higher proportion of critical incidents than the JCAHO
Sentinel Event Database. Despite coming from only 90 participating sites, AIMS received more
reports over a similar time frame than the JCAHO did from the several thousand accredited
United States hospitals. This disparity may be explained by the fact that AIMS institutions were
self-selected, and that the culture of anesthesia is more attuned to patient safety concerns. 47
The poor capture rate of incident reporting systems in American hospitals has not gone
unnoticed. Cullen et al 24 prospectively investigated usual hospital incident reporting compared to
observational data collection for adverse drug events (ADE logs, daily solicitation from hospital
personnel, and chart review) in 5 patient care units of a tertiary care hospital. Only 6% of ADEs
were identified and only 8% of serious ADEs were reported. These findings are similar to those
in the pharmacy literature, 48,49 and are attributed to cultural and environmental factors. A similar
study on a general surgical service found that 40% of patients suffered complications. 39 While
chart documentation was excellent (94%), only 20% of complications were discussed at
Morbidity and Mortality rounds.
Active solicitation of physician reporting has been suggested as a way to improve
adverse event and near miss detection rates. Weingart et al 41 employed direct physician
interviews supplemented by email reminders to increase detection of adverse events in a tertiary
care hospital. The physicians reported an entirely unique set of adverse events compared with
those captured by the hospital incident reporting system. Of 168 events, only one was reported
by both methods. O’Neil et al 37 used e-mail to elicit adverse events from housestaff and
compared these with those found on retrospective chart review. Of 174 events identified, 41
were detected by both methods. The house officers appeared to capture preventable adverse
events at a higher rate (62.5% v. 32%, p=0.003). In addition, the hospital’s risk management
system detected only 4 of 174 adverse events. Welsh et al 40 employed prompting of house
officers at morning report to augment hospital incident reporting systems. There was overlap in
reporting in only 2.6% of 341 adverse events that occurred during the study. In addition,
although the number of events house officers reported increased with daily prompting, the
quantity rapidly decreased when prompting ceased. In summary, there is evidence that active
solicitation of critical incident reports by physicians can augment existing databases, identifying
incidents not detected through other means, although the response may not be durable. 37,40,41
Potential for Harm
Users may view reporting systems with skepticism, particularly the system’s ability to
maintain confidentiality and shield participants from legal exposure. 28 In many states, critical
45

incident reporting and analysis count as peer review activities and are protected from legal
discovery. 28,50 However, other states offer little or no protection, and reporting events to external
agencies (eg, to JCAHO) may obliterate the few protections that do exist. In recognition of this
problem, JCAHO’s Terms of Agreement with hospitals now includes a provision identifying
JCAHO as a participant in each hospital’s quality improvement process. 28
Costs and Implementation
Few estimates of costs have been reported in the literature. In general, authors remarked
that incident reporting was far less expensive than retrospective review. One single center study
estimated that physician reporting was less costly ($15,000) than retrospective record review
($54,000) over a 4-month period. 37 A survey of administrators of reporting systems from nonmedical
industries reported a consensus that costs were far offset by the potential benefits. 6
Comment
The wide variation in reporting of incidents may have more to do with reporting
incentives and local culture than with the quality of medicine practiced there. 24 When institutions
prioritize incident reporting among medical staff and trainees, however, the incident reporting
systems seem to capture a distinct set of events from those captured by chart review and
traditional risk management 40,41 and events captured in this manner may be more preventable. 37
The addition of anonymous or non-punitive systems is likely to increase the rates of
incident reporting and detection. 51 Other investigators have also noted increases in reporting
when new systems are implemented and a culture conducive to reporting is maintained. 40,52
Several studies suggest that direct solicitation of physicians results in reports that are more likely
to be distinct, preventable, and more severe than those obtained by other means. 8,37,41
The nature of incident reporting, replete with hindsight bias, lost information, and lost
contextual clues makes it unlikely that robust data will ever link it directly with improved
outcomes. Nonetheless, incident reporting appears to be growing in importance in medicine. The
Institute of Medicine report, To Err is Human, 53 has prompted calls for mandatory reporting of
medical errors to continue in the United States. 54-57 England’s National Health Service plans to
launch a national incident reporting system as well, which has raised concerns similar to those
voiced in the American medical community. 58 While the literature to date does not permit an
evidence-based resolution of the debate over mandatory versus voluntary incident reporting, it is
clear that incident reporting represents just one of several potential sources of information about
patient safety and that these sources should be regarded as complementary. In other industries
incident reporting has succeeded when it is mandated by regulatory agencies or is anonymous
and voluntary on the part of reporters, and when it provides incentives and feedback to
reporters. 6 The ability of health care organizations to replicate the successes of other industries in
their use of incident reporting systems 6 will undoubtedly depend in large part on the uses to
which they put these data. Specifically, success or failure may depend on whether health care
organizations use the data to fuel institutional quality improvement rather than to generate
individual performance evaluations.
46

Table 4.1. Examples of events reported to hospital incident reporting systems 25-27
Adverse Outcomes Procedural Breakdowns Catastrophic Events
Unexpected death or
disability
Inpatient falls or
“mishaps”
Institutionally-acquired
burns
Institutionally-acquired
pressure sores
Errors or unexpected
complications related to the
administration of drugs or
transfusion
Discharges against medical
advice (“AMA”)
Significant delays in
diagnosis or diagnostic
testing
Breach of confidentiality
47
Performance of a procedure on the
wrong body part (“wrong-site
surgery”)
Performance of a procedure on the
wrong patient
Infant abduction or discharge to
wrong family
Rape of a hospitalized patient
Suicide of a hospitalized patient
Table 4.2. Sentinel event alerts published by JCAHO following analysis of incident
reports 46
• Medication Error Prevention – Potassium Chloride
• Lessons Learned: Wrong Site Surgery
• Inpatient Suicides: Recommendations for Prevention
• Preventing Restraint Deaths
• Infant Abductions: Preventing Future Occurrences
• High-Alert Medications and Patient Safety
• Operative and Postoperative Complications: Lessons for the Future
• Fatal Falls: Lessons for the Future
• Infusion Pumps: Preventing Future Adverse Events
• Lessons Learned: Fires in the Home Care Setting
• Kernicterus Threatens Healthy Newborns

References
1. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, et al. Incidence
of adverse events and negligence in hospitalized patients. Results of the Harvard Medical
Practice Study I. N Engl J Med. 1991;324:370-376.
2. Studdert D, Thomas E, Burstin H, Zbar B, Orav J, Brennan T. Negligent care and
malpractice claiming behavior in Utah and Colorado. Medical Care. 2000;38:250-260.
3. Van der Schaaf T. Near miss reporting in the Chemical Process Industry [Doctoral Thesis].
Eindhoven, The Netherlands: Eindhoven University of Technology; 1992.
4. Ibojie J, Urbaniak S. Comparing near misses with actual mistransfusion events a more
accurate reflection of transfusion errors. Br J Haematol. 2000;108:458-460.
5. Battles J, Kaplan H, Van der Schaaf T, Shea C. The attributes of medical event-reporting
systems. Arch Pathol Lab Med. 1998;122:231-238.
6. Barach P, Small S. Reporting and preventing medical mishaps: lessons from non-medical
near miss reporting systems. BMJ. 2000;320:759-763.
7. Whipple J, Quebbeman E, Lewis K, Gaughan L, Gallup E, Ausman R. Identification of
patient controlled analgesia overdoses in hospitalized patients: a computerized method of
monitoring adverse events. Ann Pharmacother. 1994;28:655-658.
8. Bates D, Makary M, Teich J, Pedraza L, Ma'luf N, Burstin H, et al. Asking residents about
adverse events in a computer dialogue: how accurate are they? Jt Comm J Qual Improv.
1998;24:197-202.
9. Jha A, GJ K, Teich J. Identifying adverse drug events: development of a computer-based
monitor and comparison with chart review and stimulated voluntary report. J Am Med
Inform Assoc. 1998;5:305-314.
10. Naranjo C, Lanctot K. Recent developments in computer-assisted diagnosis of putative
adverse drug reactions. Drug Saf. 1991;6:315-322.
11. Classen D, Pestotnik S, Evans R, Burke J. Computerized surveillance of adverse drug
events in hospital patients. JAMA. 1991;266:2847-2851.
12. Flanagan J. The critical incident technique. Psychol Bull. 1954;51:327-358.
13. Webb R, Currie M, Morgan C, Williamson J, Mackay P, Russell W, et al. The Australian
Incident Monitoring Study: an analysis of 2000 incident reports. Anaesth Intens Care.
1993;21:520-528.
14. Runciman W, Sellen A, Webb R, Williamson J, Currie M, Morgan C, et al. Errors,
incidents and accidents in anaesthetic practice. Anaesth Intens Care. 1993;21:506-519.
15. Fischoff B. Hindsight does not equal foresight: the effect of outcome knowledge on
judgment under uncertainty. J Exp Psych: Human Perform and Percept. 1975;1:288-299.
16. Food and Drug Administration: Biological products; reporting of errors and accidents in
manufacturing. Federal Register. 1997;62:49642-49648.
17. Safren MA, Chapanis A. A critical incident study of hospital medication errors. Hospitals.
1960;34:32-34.
18. Graham JR, Winslow WW, Hickey MA. The critical incident technique applied to
postgraduate training in psychiatry. Can Psychiatr Assoc J. 1972;17:177-181.
19. Cooper JB, Newbower RS, Long CD, McPeek B. Preventable anesthesia mishaps: a study
of human factors. Anesthesiology. 1978;49:399-406.
20. Buckley T, Short T, Rowbottom Y, OH T. Critical incident reporting in the intensive care
unit. Anaesthesia. 1997;52:403-409.
48

21. Hart G, Baldwin I, Gutteridge G, Ford J. Adverse incident reporting in intensive care.
Anaesth Intens Care. 1994;22:556-561.
22. Linden JV, Paul B, Dressler KP. A report of 104 transfusion errors in New York State.
Transfusion. 1992;32:601-606.
23. Emori T, Edwards J, Culver D, Sartor C, Stroud L, Gaunt E, et al. Accuracy of reporting
nosocomial infections in intensive-care-unit patients to the National Nosocomial Infections
Surveillance System: a pilot study. Infect Control Hosp Epidemiol. 1998;19:308-316.
24. Cullen D, Bates D, Small S, Cooper J, Nemeskal A, Leape L. The incident reporting
system does not detect adverse events: a problem for quality improvement. Jt Comm J
Qual Improv. 1995;21:541-548.
25. Incident reports. A guide to legal issues in health care. Philadelphia, PA: Trustees of the
University of Pennsylvania. 1998:127-129.
26. Joint Commission on the Accreditation of Healthcare Organizations. Sentinel event policy
and procedures. Available at: http://JCAHO.org/sentinel/se_pp.html. Accessed March 15,
2001.
27. Fischer G, Fetters M, Munro A, Goldman E. Adverse events in primary care identified
from a risk-management database. J Fam Pract. 1997;45:40-46.
28. Berman S. Identifying and addressing sentinel events: an Interview with Richard Croteau.
Jt Comm J Qual Improv. 1998;24:426-434.
29. Joint Commission on Accreditation of Healthcare Organizations. Sentinel Event Statistics.
Available at: http://www.JCAHO.org/sentinel/se_stats.html. Accessed April 16, 2001.
30. Joint Commission on Accreditation of Healthcare Organizations. Sentinel Event Alert.
Available at: http://JCAHO.org/edu_pub/sealert/se_alert.html. Accessed May 14, 2001.
31. Lessons learned: sentinel event trends in wrong-site surgery. Joint Commission
Perspectives. 2000;20:14.
32. Symposium: The Australian Incident Monitoring Study. Anaesth Intens Care.
1993;21:506-695.
33. Billings C, Reynard W. Human factors in aircraft incidents: results of a 7-year study. Aviat
Space Environ Med. 1984;55:960-965.
34. National Aeronautics & Space Agency (NASA). Aviation Safety Reporting System.
Available at: http://www-afo.arc.nasa.gov/ASRS/ASRS.html. Accessed March 14, 2000.
35. Leape L. Reporting of medical errors: time for a reality check. Qual Health Care.
2000;9:144-145.
36. New York State Department of Health. NYPORTS - The New York Patient Occurrence
and Tracking System. Available at:
http://www.health.state.ny.us/nysdoh/commish/2001/nyports/nyports.htm. Accessed May
31, 2001.
37. O'Neil A, Petersen L, Cook E, Bates D, Lee T, Brennan T. Physician reporting compared
with medical-record review to identify adverse medical events. Ann Intern Med.
1993;119:370-376.
38. Data sources and coordination. In: Surgeons ACo, editor. Patient safety manual. Rockville,
MD: Bader & Associates, Inc. 1985.
39. Wanzel K, Jamieson C, Bohnen J. Complications on a general surgery service: incidence
and reporting. CJS. 2000;43:113-117.
40. Welsh C, Pedot R, Anderson R. Use of morning report to enhance adverse event detection.
J Gen Intern Med. 1996;11:454-460.
49

41. Weingart SN, Ship AN, Aronson MD. Confidential clinician-reported surveillance of
adverse events among medical inpatients. J Gen Intern Med. 2000;15:470-477.
42. Frey B, Kehrer B, Losa M, Braun H, Berweger L, Micallef J, et al. Comprehensive critical
incident monitoring in a neonatal-pediatric intensive care unit: experience with the system
approach. Intensive Care Med. 2000;26:69-74.
43. Findlay G, Spittal M, Radcliffe J. The recognition of critical incidents: quantification of
monitor effectiveness. Anaesthesia. 1998;53:589-603.
44. Flaatten H, Hevroy O. Errors in the intensive care unit (ICU). Act Anaesthesiol Scand.
1999;43:614-617.
45. Connolly C. Reducing error, improving safety: relation between reported mishaps and
safety is unclear. BMJ. 2000;321:505.
46. The Joint Commission on Accreditation of Healthcare Organizations. Sentinel Event Alert.
Available at: http://www.JCAHO.org/edu_pub/sealert/se_alert.html. Accessed May 31,
2001.
47. Gaba D. Anaesthesiology as a model for patient safety in health care. BMJ. 2000;320.
48. Shannon R, De Muth J. Comparison of medication error detection methods in the long term
care facility. Consulting Pharmacy. 1987;2:148-151.
49. Allan E, Barker K. Fundamentals of medication error research. Am J Hosp Pharm.
1990;47:555-571.
50. Rex J, Turnbull J, Allen S, Vande Voorde K, Luther K. Systematic root cause analysis of
adverse drug events in a tertiary referral hospital. Jt Comm J Qual Improv. 2000;26:563-
575.
51. Kaplan H, Battles J, Van Der Schaaf T, Shea C, Mercer S. Identification and classification
of the causes of events in transfusion medicine. Transfusion. 1998;38:1071-1081.
52. Hartwig S, Denger S, Schneider P. Severity-indexed, incident report-based medication
error-reporting program. Am J Hosp Pharm. 1991;48:2611-2616.
53. Kohn L, Corrigan J, Donaldson M, editors. To Err Is Human: Building a Safer Health
System. Washington, DC: Committee on Quality of Health Care in America, Institute of
Medicine. National Academy Press. 2000.
54. Pear R. U.S. health officials reject plan to report medical mistakes. New York Times Jan 24
2000: A14(N), A14(L).
55. Prager LO. Mandatory reports cloud error plan: supporters are concerned that the merits of
a plan to reduce medical errors by 50% are being obscured by debate over its most
controversial component. American Medical News March 13 2000:1,66-67.
56. Murray S. Clinton to call on all states to adopt systems for reporting medical errors. Wall
Street Journal Feb 22 2000:A28(W), A6(E).
57. Kaufman M. Clinton Seeks Medical Error Reports; Proposal to Reduce Mistakes Includes
Mandatory Disclosure, Lawsuit Shield. Washington Post Feb 22 2000:A02.
58. Mayor S. English NHS to set up new reporting system for errors. BMJ. 2000;320:1689.
50

Chapter 5. Root Cause Analysis
Heidi Wald, MD
University of Pennsylvania School of Medicine
Kaveh G. Shojania, MD
University of California, San Francisco School of Medicine
Background
Historically, medicine has relied heavily on quantitative approaches for quality
improvement and error reduction. For instance, the US Food and Drug Administration (FDA)
has collected data on major transfusion errors since the mid-1970s. 1,2 Using the statistical power
of these nationwide data, the most common types of errors have been periodically reviewed and
systems improvements recommended. 3
These epidemiologic techniques are suited to complications that occur with reasonable
frequency, but not for rare (but nonetheless important) errors. Outside of medicine, high-risk
industries have developed techniques to address major accidents. Clearly the nuclear power
industry cannot wait for several Three Mile Island-type events to occur in order to conduct valid
analyses to determine the likely causes.
A retrospective approach to error analysis, called root cause analysis (RCA), is widely
applied to investigate major industrial accidents. 4 RCA has its foundations in industrial
psychology and human factors engineering. Many experts have championed it for the
investigation of sentinel events in medicine. 5-7 In 1997, the Joint Commission on the
Accreditation of Healthcare Organizations (JCAHO) mandated the use of RCA in the
investigation of sentinel events in accredited hospitals. 8
The most commonly cited taxonomy of human error in the medical literature is based on
the work of James Reason. 4,9,10 Reason describes 2 major categories of error: active error, which
generally occurs at the point of human interface with a complex system, and latent error, which
represents failures of system design. RCA is generally employed to uncover latent errors
underlying a sentinel event. 6,7
RCA provides a structured and process-focused framework with which to approach
sentinel event analysis. Its cardinal tenet is to avoid the pervasive and counterproductive culture
of individual blame. 11,12 Systems and organizational issues can be identified and addressed, and
active errors are acknowledged. 6 Systematic application of RCA may uncover common root
causes that link a disparate collection of accidents (ie, a variety of serious adverse events
occurring at shift change). Careful analysis may suggest system changes designed to prevent
future incidents. 13
Despite these intriguing qualities, RCA has significant methodologic limitations. RCAs
are in essence uncontrolled case studies. As the occurrence of accidents is highly unpredictable,
it is impossible to know if the root cause established by the analysis is the cause of the
accident. 14 In addition, RCAs may be tainted by hindsight bias. 4,15,16 Other biases stem from how
deeply the causes are probed and influenced by the prevailing concerns of the day. 16,17 The fact
that technological failures (device malfunction), which previously represented the focus of most
accident analyses, have been supplanted by staffing issues, management failures, and
information systems problems may be an example of the latter bias. 17 Finally, RCAs are timeconsuming
and labor intensive.
51

Despite legitimate concerns about the place of RCA in medical error reduction, the
JCAHO mandate ensures that RCA will be widely used to analyze sentinel events. 8 Qualitative
methods such as RCA should be used to supplement quantitative methods, to generate new
hypotheses, and to examine events not amenable to quantitative methods (for example, those that
occur rarely). 18 As such, its credibility as a research tool should be judged by the standards
appropriate for qualitative research, not quantitative. 19,20 Yet, the outcomes and costs associated
with RCA are largely unreported. This chapter reviews the small body of published literature
regarding the use of RCA in the investigation of medical errors.
Practice Description
To be credible, RCA requires rigorous application of established qualitative techniques.
Once a sentinel event has been identified for analysis (eg, a major chemotherapy dosing error, a
case of wrong-site surgery, or major ABO incompatible transfusion reaction), a multidisciplinary
team is assembled to direct the investigation. The members of this team should be trained in the
techniques and goals of RCA, as the tendency to revert to personal biases is strong. 13,14 Multiple
investigators allow triangulation or corroboration of major findings and increase the validity of
the final results. 19 Based on the concepts of active and latent error described above, accident
analysis is generally broken down into the following steps 6,7 :
1. Data collection: establishment of what happened through structured
interviews, document review, and/or field observation. These data are used to
generate a sequence or timeline of events preceeding and following the event;
2. Data analysis: an iterative process to examine the sequence of events
generated above with the goals of determining the common underlying
factors:
(i) Establishment of how the event happened by identification of active
failures in the sequence;
(ii) Establishment of why the event happened through identification of latent
failures in the sequence which are generalizable.
In order to ensure consideration of all potential root causes of error, one popular
conceptual framework for contributing factors has been proposed based on work by Reason. 7
Several other frameworks also exist. 21,22 The categories of factors influencing clinical practice
include institutional/regulatory, organizational/management, work environment, team factors,
staff factors, task factors, and patient characteristics. Each category can be expanded to provide
more detail. A credible RCA considers root causes in all categories before rejecting a factor or
category of factors as non-contributory. A standardized template in the form of a tree (or
“Ishikawa”) may help direct the process of identifying contributing factors, with such factors
leading to the event grouped (on tree “roots”) by category. Category labels may vary depending
on the setting. 23
At the conclusion of the RCA, the team summarizes the underlying causes and their
relative contributions, and begins to identify administrative and systems problems that might be
candidates for redesign. 6
Prevalence and Severity of the Target Safety Problem
JCAHO’s 6-year-old sentinel event database of voluntarily reported incidents (see
Chapter 4) has captured a mere 1152 events, of which 62% occurred in general hospitals. Twothirds
of the events were self-reported by institutions, with the remainder coming from patient
52

complaints, media stories and other sources. 24 These statistics are clearly affected by
underreporting and consist primarily of serious adverse events (76% of events reported resulted
in patient deaths), not near misses. The number of sentinel events appropriate for RCA is likely
to be orders of magnitude greater.
The selection of events for RCA may be crucial to its successful implementation on a
regular basis. Clearly, it cannot be performed for every medical error. JCAHO provides guidance
for hospitals about which events are considered “sentinel,” 8 but the decision to conduct RCA is
at the discretion of the leadership of the organization. 12
If the number of events is large and homogeneous, many events can be excluded from
analysis. In a transfusion medicine reporting system, all events were screened after initial report
and entered in the database, but those not considered sufficiently unique did not undergo RCA. 25
Opportunities for Impact
While routine RCA of sentinel events is mandated, the degree to which hospitals carry
out credible RCAs is unknown. Given the numerous demands on hospital administrators and
clinical staff, it is likely that many hospitals fail to give this process a high profile, assigning the
task to a few personnel with minimal training in RCA rather than involving trained leaders from
all relevant departments. The degree of underreporting to JCAHO suggests that many hospitals
are wary of probationary status and the legal implications of disclosure of sentinel events and the
results of RCAs. 12,26
Study Designs
As RCA is a qualitative technique, most reports in the literature are case studies or case
series of its application in medicine. 6,27-30 There is little published literature that systematically
evaluates the impact of formal RCA on error rates. The most rigorous study comes from a
tertiary referral hospital in Texas that systematically applied RCA to all serious adverse drug
events (ADEs) considered preventable. The time series contained background data during the
initial implementation period of 12 months and a 17-month follow-up phase. 13
Study Outcomes
Published reports of the application of RCA in medicine generally present incident
reporting rates, categories of active errors determined by the RCA, categories of root causes
(latent errors) of the events, and suggested systems improvements. While these do not represent
clinical outcomes, they are reasonable surrogates for evaluation. For instance, increased incident
reporting rates may reflect an institution’s shift toward increased acceptance of quality
improvement and organizational change. 5,21
53

Evidence for Effectiveness of the Practice
The Texas study revealed a 45% decrease in the rate of voluntarily reported serious
ADEs between the study and follow-up periods (7.2 per 100,000 to 4.0 per 100,000 patient-days,
p

investigation in the name of quality improvement, with one health system reporting a sustained
10-fold increase in reporting. 25,27
Comment
Root cause analyses systematically search out latent or system failures that underlie
adverse events or near misses. They are limited by their retrospective and inherently speculative
nature. There is insufficient evidence in the medical literature to support RCA as a proven
patient safety practice, however it may represent an important qualitative tool that is
complementary to other techniques employed in error reduction. When applied appropriately,
RCA may illuminate targets for change, and, in certain health care contexts, may generate
testable hypotheses. The use of RCA merits more consideration, as it lends a formal structure to
efforts to learn from past mistakes.
References
1. Sazama K. Current good manufacturing practices for transfusion medicine. Transfus Med
Rev. 1996;10:286-295.
2. Food and Drug Administration: Biological products; reporting of errors and accidents in
manufacturing. Fed Regist. 1997;62:49642-49648.
3. Sazama K. Reports of 355 transfusion-associated deaths: 1976 through 1985. Transfusion.
1990;30:583-590.
4. Reason JT. Human Error. New York: Cambridge Univ Press. 1990.
5. Battles JB, Kaplan HS, Van der Schaaf TW, Shea CE. The attributes of medical eventreporting
systems: experience with a prototype medical event-reporting system for
transfusion medicine. Arch Pathol Lab Med. 1998;122:231-238.
6. Eagle CJ, Davies JM, Reason J. Accident analysis of large-scale technological disasters
applied to an anaesthetic complication. Can J Anaesth. 1992;39:118-122.
7. Vincent C, Ennis M, Audley RJ. Medical accidents. Oxford ; New York: Oxford
University Press. 1993.
8. Joint Commission on the Accreditation of Healthcare Organizations. Sentinel event policy
and procedures. Available at: http://www.JCAHO.org/sentinel/se_pp.html. Accessed May
30, 2001.
9. Reason JT. Managing the Risks of Organizational Accidents: Ashgate Publishing
Company. 1997.
10. Reason J. Human error: models and management. BMJ. 2000;320:768-770.
11. Leape LL. Error in medicine. JAMA. 1994;272:1851-1857.
12. Berman S. Identifying and addressing sentinel events: an Interview with Richard Croteau.
Jt Comm J Qual Improv. 1998;24:426-434.
13. Rex JH, Turnbull JE, Allen SJ, Vande Voorde K, Luther K. Systematic root cause analysis
of adverse drug events in a tertiary referral hospital. Jt Comm J Qual Improv. 2000;26:563-
575.
14. Runciman WB, Sellen A, Webb RK, Williamson JA, Currie M, Morgan C, et al. The
Australian Incident Monitoring Study. Errors, incidents and accidents in anaesthetic
practice. Anaesth Intensive Care. 1993;21:506-519.
15. Caplan RA, Posner KL, Cheney FW. Effect of outcome on physician judgments of
appropriateness of care. JAMA. 1991;265:1957-1960.
16. Perrow C. Normal accidents : Living with High-Risk Technologies. With a New Afterword
and a Postscript on the Y2K Problem. Princeton, NJ: Princeton University Press. 1999.
55

17. Rasmussen J. Human error and the problem of causality in analysis of accidents. Philos
Trans R Soc Lond B Biol Sci. 1990;327:449-460.
18. Pope C, Mays N. Reaching the parts other methods cannot reach: an introduction to
qualitative methods in health and health services research. BMJ. 1995;311:42-45.
19. Giacomini MK, Cook DJ. Users' guides to the medical literature: XXIII. Qualitative
research in health care A. Are the results of the study valid? Evidence-Based Medicine
Working Group. JAMA. 2000;284:357-362.
20. Giacomini MK, Cook DJ. Users' guides to the medical literature: XXIII. Qualitative
research in health care B. What are the results and how do they help me care for my
patients? Evidence-Based Medicine Working Group. JAMA. 2000;284:478-482.
21. Van der Schaaf T. Near miss reporting in the Chemical Process Industry [Doctoral thesis].
Eindhoven, The Netherlands: Eindhoven University of Technology. 1992.
22. Moray N. Error reduction as a systems problem. In: Bogner M, editor. Human error in
medicine. Hillsdale, NJ: Lawrence Erlbaum Associates. 1994.
23. Ishikawa K. What is total quality control? The Japanese way. Englewood Cliffs, NJ:
Prentice Hall. 1985.
24. Joint Commission on Accreditation of Healthcare Organizations. Sentinel Event Statistics.
Available at: http://www.JCAHO.org. Accessed April 16, 2001.
25. Kaplan HS, Battles JB, Van der Schaaf TW, Shea CE, Mercer SQ. Identification and
classification of the causes of events in transfusion medicine. Transfusion. 1998;38:1071-
1081.
26. Will hospitals come clean for JCAHO? Trustee. 1998;51:6-7.
27. Leape LL, Bates DW, Cullen DJ, Cooper J, Demonaco HJ, Gallivan T, et al. Systems
analysis of adverse drug events. ADE Prevention Study Group. JAMA. 1995;274:35-43.
28. Bhasale A. The wrong diagnosis: identifying causes of potentially adverse events in
general practice using incident monitoring. Fam Pract. 1998;15:308-318.
29. Troidl H. Disasters of endoscopic surgery and how to avoid them: error analysis. World J
Surg. 1999;23:846-855.
30. Fernandes CM, Walker R, Price A, Marsden J, Haley L. Root cause analysis of laboratory
delays to an emergency department. J Emerg Med. 1997;15:735-739.
31. Hofer TP, Kerr EA. What is an error? Eff Clin Pract. 2000;3:261-269.
56

PART III. PATIENT SAFETY PRACTICES & TARGETS
Section A. Adverse Drug Events (ADEs)
Section B. Infection Control
Section C. Surgery, Anesthesia, and Perioperative Medicine
Section D. Safety Practices for Hospitalized or Institutionalized Elders
Section E. General Clinical Topics
Section F. Organization, Structure, and Culture
Section G. Systems Issues and Human Factors
Section H. Role of the Patient
57

Section A. Adverse Drug Events (ADEs)
Chapter 6. Computerized Physician Order Entry (CPOE) with Clinical Decision Support
Systems (CDSSs)
Chapter 7. The Clinical Pharmacist’s Role in Preventing Adverse Drug Events
Chapter 8. Computer Adverse Drug Event (ADE) Detection and Alerts
Chapter 9. Protocols for High-Risk Drugs: Reducing Adverse Drug Events Related to
Anticoagulants
Chapter 10. Unit-Dose Drug Distribution Systems
Chapter 11. Automated Medication Dispensing Devices
58

Chapter 6. Computerized Physician Order Entry (CPOE) with Clinical
Decision Support Systems (CDSSs)
Rainu Kaushal, MD, MPH
David W. Bates, MD, MSc
Harvard Medical School
Background
Medication errors and adverse drug events (ADEs) are common, costly, and clinically
important problems. 1-7 Two inpatient studies, one in adults and one in pediatrics, have found
that about half of medication errors occur at the stage of drug ordering, 2, 7 although direct
observation studies have indicated that many errors also occur at the administration stage. 8 The
principal types of medication errors, apart from missing a dose, include incorrect medication
dose, frequency, or route. 2 ADEs are injuries that result from the use of a drug. Systems-based
analysis of medication errors and ADEs suggest that changes in the medication ordering system,
including the introduction of computerized physician order entry (CPOE) with clinical decision
support systems (CDSSs), may reduce medication-related errors. 9
Despite growing evidence and public mandates, implementation of CPOE has been
limited. The Leapfrog Group, a consortium of companies from the Business Roundtable, has
endorsed CPOE in hospitals as one of the 3 changes that would most improve patient safety in
America (see also Chapter 55). 10 A Medicare Payment Advisory Commission report suggested
instituting financial incentives for CPOE implementation. 11 US Senators Bob Graham (D-Fla.)
and Olympia Snowe (R-Maine) recently introduced a bill, entitled the “Medication Errors
Reduction Act of 2001,” to establish an informatics system grant program for hospitals and
skilled nursing facilities. 12 In addition, California recently enacted legislation stipulating that
acute care hospitals implement information technology, such as CPOE to reduce medicationrelated
errors. 13
Practice Description
CPOE refers to a variety of computer-based systems of ordering medications, which
share the common features of automating the medication ordering process. Basic CPOE ensures
standardized, legible, complete orders by only accepting typed orders in a standard and complete
format. Almost all CPOE systems include or interface with CDSSs of varying sophistication.
Basic clinical decision support may include suggestions or default values for drug doses, routes,
and frequencies. More sophisticated CDSSs can perform drug allergy checks, drug-laboratory
value checks, drug-drug interaction checks, in addition to providing reminders about corollary
orders (eg, prompting the user to order glucose checks after ordering insulin) or drug guidelines
to the physician at the time of drug ordering (see also Chapter 53).
At times, CDSSs are implemented without CPOE. Isolated CDSSs can provide advice on
drug selection, dosages, and duration. More refined CDSSs can incorporate patient-specific
information (for example recommending appropriate anticoagulation regimens), or incorporate
pathogen-specific information such as suggesting appropriate anti-infective regimens. After
viewing such advice, the physician proceeds with a conventional handwritten medication order.
59

Prevalence and Severity of the Target Safety Problem
It is estimated that over 770,000 people are injured or die in hospitals from ADEs
annually. 4, 5, 14 The few hospitals that have studied incidence rates of ADEs have documented
rates ranging from 2 to 7 per 100 admissions. 2, 4, 15, 16 A precise national estimate is difficult to
calculate due to the variety of criteria and definitions used by researchers. 17 One study of
preventable inpatient ADEs in adults demonstrated that 56% occurred at the stage of ordering,
34% at administration, 6% at transcribing, and 4% at dispensing. 2 In this study, the drug class
most commonly associated with preventable ADEs was analgesics, followed by sedatives and
antibiotics. Even fewer studies have been conducted in the outpatient setting. One recent crosssectional
chart review and patient care survey found an ADE rate of 3% in adult primary care
outpatients. 18
Opportunities for Impact
Clear data do not exist about the prevalence of CPOE with CDSSs or isolated CDSSs.
One survey of 668 hospitals indicated that 15% had at least partially implemented CPOE. 19 A
slightly more recent survey of pharmacy directors at 1050 acute care hospitals in the United
States (51% response rate) reported that 13% of hospitals had an electronic medication orderentry
system in place, while an additional 27% stated they were in the process of obtaining such
a sysytem. 20
Study Designs
The 4 studies listed in Table 6.1 evaluated CPOE with CDSSs. 21-24 The first study, a
randomized controlled trial evaluating the utility of CPOE in improving prescribing for corollary
orders, was conducted by the Regenstrief Institute for Health Care (affiliated with the Indiana
University School of Medicine). 21 The remaining 3 studies evaluated the CPOE system at
Brigham and Women’s Hospital (BWH). 22-24 Of note, both the Regenstrief and BWH systems
are “home-grown” rather than “off-the-shelf” commercial systems. The first BWH study was a
cross-sectional analysis comparing an intervention period of CPOE with CDSSs for adult
inpatients on medical, surgical, and intensive care wards with a historical period. 22 The other 2
23, 24
BWH studies were time series analyses of orders written for adult inpatients.
Table 6.2 lists 4 studies 25-28 that evaluated isolated CDSSs, 2 of which represent
systematic reviews. 25, 26 Hunt et al 25 updated an earlier systematic review, 29 and included 68
studies that prospectively evaluated use of CDSSs in a clinical setting by a healthcare
practitioner with a concurrent control. Importantly, Hunt et al included studies of outpatient
settings in their review. Walton’s review addressed 15 articles that studied computerized advice
on drug dosage for inpatients. 26 Evans et al, at Latter Day Saints Hospital, performed the other 2
studies, 27, 28 again on a “home-grown” system. The first was a randomized controlled trial of
empiric antibiotic selection using CDSSs. 27 The second study was a cross-sectional analysis
comparing an intervention period of a computer-assisted management program for antiinfectives
with a historical control period. 28 This second study 28 was likely excluded from the
systematic review by Hunt et al 25 for methodologic reasons, as it did not have a concurrent
control. The reasons for excluding the first study 27 remain unclear. Finally, it is important to
emphasize again that of the primary studies that were included, 6 of 8 were performed at 3
institutions with sophisticated “home-grown” systems.
60

Study Outcomes
Adverse drug events (ADEs), (injuries that result from the use of drugs) by definition
constitute clinical outcomes (Level 1). ADEs that are associated with a medication error are
considered preventable, while those not associated with a medication error (eg, known
medication side effects) are considered non-preventable. An example of a preventable ADE is
the development of rash after the administration of ampicillin to a known penicillin-allergic
patient. In contrast, a non-preventable ADE would be development of an ampicillin-associated
rash in a patient with no known drug allergies. Non-intercepted serious medication errors include
non-intercepted potential ADEs and preventable ADEs (ie, medication errors that either have the
potential or actually cause harm to a patient).
Medication errors refer to errors in the processes of ordering, transcribing, dispensing,
administering, or monitoring medications, irrespective of the outcome (ie, injury to the patient).
One example is an order written for amoxicillin without a route of administration. Other
medication errors have a greater potential for patient harm and so are often designated as
“serious medication errors” or “potential ADEs” – eg, an order for amoxicillin in a patient with
past anaphylaxis to penicillin.
Potential ADEs may or may not be intercepted before reaching the patient. An example
of an intercepted potential ADE would be an order written for an acetaminophen overdose that is
intercepted and corrected by a nurse before reaching the patient. An example of a nonintercepted
potential ADE would be an administered overdose of acetaminophen to a patient who
did not suffer any sequelae.
Medication errors include a mixture of serious medication errors with a significant
potential for patient injury (Level 2) and other deviations from recommended practice that do not
have a clear or established connection to adverse events (Level 3). Examples of the latter
include failure to choose the optimal dosing schedule for a medication and many standards
related to monitoring serum drug levels and routine electrolytes.
Only 2 studies (from a single institution) evaluating CPOE with CDSSs included ADEs
as a secondary outcome (Level 1), 22 and even these studies primary outcomes were serious
medication errors (Level 2) and non-intercepted medication errors. 23 The other studies reported
a variety of other errors often involving mixtures of Level 2 and Level 3 outcomes – eg,
“prescribing practices” 24 and “corollary orders.” 21 Corollary orders refer to orders required to
detect or ameliorate adverse reactions that may result from the trigger order - eg, checking the
serum creatinine a minimum of 2 times per week in a patient receiving a nephrotoxic agent such
as amphotericin. Many corollary orders capture Level 3 outcomes, as failure to prescribe these
orders would in most cases have no clear impact on clinical outcomes – eg, failure to order daily
tests for fecal occult blood in patients on heparin or screening audiometry for patients receiving
vancomycin. 21
The predominance of Level 2 and 3 outcomes in these studies is understandable, given
the much higher frequency of these outcomes compared to actual ADEs and the enormous costs
of conducting these studies.
Similarly, the studies evaluating CDSSs reported a mixture of outcomes from Levels 1-3.
Hunt et al reviewed articles to determine changes in patient outcomes (Level 1) or physician
performance (Levels 2 and 3, depending on the practice). 25 Walton et al evaluated a range of
outcomes (Levels 1-3), including reductions in “adverse reactions and unwanted effects” (Level
1). 26 In one study, Evans et al determined rates of pathogen susceptibility to an antibiotic
61

egimen 27 (Level 2); another study by the same authors evaluated adverse drug events caused by
anti-infectives as a main outcome (Level 1). 28
Evidence for Effectiveness of the Practice
Of the 2 studies at BWH that addressed the impact of CPOE with CDSSs on medication
errors and ADEs, the first demonstrated a 55% decrease in serious medication errors. 22 As a
secondary outcome, this study found a 17% decrease in preventable ADEs, which was not
statistically significant. The second study, a time series analysis, demonstrated marked
reductions in all medication errors excluding missed dose errors and in non-intercepted serious
medication errors. 23 The number of ADEs in this study was quite small – 25 in the baseline
period and 18 in the third of the 3 periods with CPOE and CDSS. Correcting for the number of
opportunities for errors, the total number of ADEs/1000 patient days decreased from 14.7 to 9.6
(p=0.09). For the sub-category of preventable ADEs, the reduction (from 5 to 2) achieved
borderline statistical significance (p=0.05).
Overhage et al and Teich et al studied more focused aspects of the medication system.
Overhage et al 21 studied computerized reminders for corollary orders (eg, entering a laboratory
order to check electrolytes when ordering potassium for a patient) and showed a greater than
100% improvement in the rate of corollary orders (p

Potential for Harm
Faulty decision support data, for example an incorrect default dosing suggestion, can lead
to inappropriate ordering choices by physicians. The BWH time series analysis demonstrated an
initial rise in intercepted potential ADEs due to the structure of the ordering screen for potassium
chloride. 23 This structural error was identified and easily rectified, but underscores the
importance of close scrutiny of all aspects of CPOE screens, both initially and on an ongoing
basis. 23
Also, analogously to writing an order in the wrong patient chart in a conventional system,
a physician can electronically write an order in the wrong patient’s record - eg, after walking
away from the terminal, opening the wrong record from a personalized rounding list. In
addition, it is critical that the trigger level for warnings appropriately balances alarm sensitivity
and specificity. These systems must have thresholds set so that physicians receive warnings in
situations with a potential for significant harm, without being overwhelmed by “false alarms.”
Another potential risk is hardware outage or software instability. For example, the reliability
that is needed with CPOE is much higher than that required for systems that simply report
laboratory results.
Costs and Implementation
Six of the studies described in this review evaluated “home-grown” rather than “off-theshelf”
systems. The present costs for purchasing commercial systems are substantially more
than the previous costs of developing such systems. For BWH, the cost of developing and
implementing CPOE in 1992 was estimated to be $1.9 million, with maintenance costs of
$500,000 per year. In comparison, the cost of purchasing and implementing large commercial
systems varies substantially, but may be on the order of tens of millions of dollars. Several
studies demonstrate that only minimal resources are needed to introduce and/or maintain
21, 24, 30
decision support programs into existing order entry programs.
Relatively few data are available regarding the financial benefits of CPOE, although they
extend well beyond medication-related events. The net savings of the BWH system are
estimated at $5-10 million per year. 31 It is estimated that the costs to BWH for preventable
ADEs are $2.8 million annually. 32 Evans et al reported a $100,000 per year cost avoidance with
a computer-assisted antibiotic dosing program largely attributable to decreased antibiotic use and
avoided ADEs. 33
Importantly, healthcare systems must garner both financial and organizational support
before introducing CPOE with CDSSs. CPOE requires a very large up-front investment with
more remote, albeit substantial returns. In addition, CPOE impacts clinicians and workflow
substantially. Its complexity requires close integration with multiple systems, such as the
laboratory and pharmacy systems. Failure to attend to the impact of such a large-scale effort on
organizational culture and dynamics may result in implementation failure. 34 Therefore, it is
essential to have organizational support and integration for its successful implementation and
use.
Comment
The literature supports CPOE’s beneficial effect in reducing the frequency of a range of
medication errors, including serious errors with the potential for harm. Fewer data are available
regarding the impact of CPOE on ADEs, with no study showing a significant decrease in actual
patient harm. Similarly, isolated CDSSs appear to prevent a range of medication errors, but with
63

few data describing reductions in ADEs or improvements in other clinical outcomes. Finally, the
studied CDSSs address focused medication use (for example, antibiotic dosing) rather than more
general aspects of medication use.
Further research should be conducted to compare the various types of systems and to
compare “home-grown” with commercially available systems. Such comparisons are
particularly important since the institutions that have published CPOE outcomes have generally
been those with strong institutional commitments to their systems. Whether less committed
institutions purchasing “off the shelf” systems will see benefits comparable to those enjoyed by
“pioneers” with home-grown systems remains to be determined. Studying the benefits of such
complex systems requires rigorous methodology and sufficient size to provide the power to
study ADEs. Further research also needs to address optimal ways for institutions to acquire and
implement computerized ordering systems.
64

Table 6.1. Studies of computerized physician order entry (CPOE) with clinical decision
support systems (CDSSs) *
Study Study Design Study Outcomes Results
Overhage, 1997. 21 Impact of
faculty and physician
reminders (using CPOE) on
corollary orders for adult
inpatients in a general medical
ward at a public teaching
hospital affiliated with the
Indiana University School of
Medicine
Bates, 1998. 22 CPOE with
CDSSs for adult inpatients on
medical, surgical, and intensive
care wards at BWH, a tertiary
care center affiliated with
Harvard University
Bates, 1999. 23 CPOE with
CDSSs for adult inpatients in 3
medical units at BWH
Teich, 2000. 24 CPOE with
CDSSs for all adult inpatients at
BWH
Level 1
(RCT with
physicians
randomized
to receive
reminders or
not)
Levels 2 & 3
(two study
designs)
Level 3
(retrospective
time series)
Level 3
(retrospective
before-after
analysis)
65
Levels 2 & 3
(errors of omission
in corollary orders)
Level 1 (ADE
rates) and Level 2
(serious
medication errors)
Level 1 (ADEs)
and Level 2 (main
outcome measure
was medication
errors)
Levels 2 & 3
(changes in 5
prescribing
practices)
25% improvement in
ordering of corollary
medications by faculty
and residents
(p

Table 6.2. Studies of clinical decision support systems (CDSSs) *
Study Setting Study Design Study Outcomes Results
Hunt, 1998. 25 Use of
CDSSs by healthcare
practitioners in
multiple inpatient and
outpatient settings
Walton, 2001. 26 Use of
CDSSs for drug dosage
advice by healthcare
practitioners for 1229
patients in multiple
inpatient settings
Evans, 1994. 27 Use of a
computerized antibiotic
selection consultant for
451 inpatients at Salt
Lake City’s LDS
Hospital, a 520-bed
community teaching
hospital and tertiary
referral center
Evans, 1998. 28
Computer-based antiinfective
management
program for 1136
intensive care unit
patients from a 12-bed
ICU at LDS Hospital
Level 1A
(systematic
review of RCTs)
Levels 1A-3A
(systematic
review of RCTs,
interrupted time
series analyses,
and controlled
before-after
studies)
Level 1 (RCT
with crossover
design)
Level 2
(prospective
before-after
analysis)
Levels 1 & 2 (a
variety of measures
related to patient
outcomes and
physician practice, not
just ADEs and
processes of care
related to medication
use.
Level 1 (one main
outcome measure was
adverse drug reactions
Level 2 (one of 5
primary outcomes was
pathogen
susceptibility to
prescribed antibiotic
regimens
Level 2 (one primary
outcome was ADEs
due to anti-infective
agents
66
6 of 14 studies showed
improvement in patient
outcomes
43 of 65 studies
showed improvement
in physician
performance
Absolute risk reduction
with CDSSs: 6% (95%
CI: 0-12%)
17% greater pathogen
susceptibility to an
antibiotic regimen
suggested by computer
consultant versus
physicians (p

References
1. Barker KN, Mikeal RL, Pearson RE, Illig NA, Morse ML. Medication errors in nursing
homes and small hospitals. Am J Hosp Pharm. 1982; 39: 987-91.
2. Bates DW, Cullen DJ, Laird N, Petersen LA, Small SD, Servi D, et al. Incidence of
adverse drug events and potential adverse drug events: Implications for prevention.
JAMA. 1995; 274: 29-34.
3. Dean BS, Allan EL, Barber ND, Barker KN. Comparison of medication errors in an
American and British hospital. Am J Health Syst Pharm. 1995; 52: 2543-49.
4. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in
hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA.
1997; 277: 301-6.
5. Cullen DJ, Sweitzer BJ, Bates DW, Burdick E, Edmondson A, Leape LL. Preventable
adverse drug events in hospitalized patients: a comparative study of intensive care and
general care units. Crit Care Med. 1997; 25: 1289-97.
6. Lesar TS, Lomaestro BM, Pohl H. Medication-prescribing errors in a teaching hospital.
A 9-year experience. Arch Intern Med. 1997; 157: 1569-76.
7. Kaushal R, Bates DW, Landrigan C, McKenna J, Clapp MD, Federico F, et al.
Medication Errors and Adverse Drug Events in Pediatric Inpatients. JAMA. 2001; 285:
2114-20.
8. Allan EL, Barker KN. Fundamentals of medication error research. Am J Hosp Pharm.
1990; 47: 555-71.
9. Leape LL, Bates DW, Cullen DJ, Cooper J, Demonaco HJ, Gallivan T, et al. Systems
analysis of adverse drug events. ADE Prevention Study Group. JAMA. 1995; 274: 35-43.
10. Milstein A, Galvin RS, Delbanco SF, Salber P, Buck Jr CR. Improving the safety of
health care: the leapfrog initiative. Eff Clin Pract. 2000; 3: 313-6.
11. The Medicare Payment Advisory Commission. Report to the Congress. Selected
Medicare Issues. Available at: http://www.medpac.gov/. Accessed May 21, 2001.
12. United States Senate. Press Release. Available at:
http://www.senate.gov/~graham/pr050301.html. Accessed June 12, 2001.
13. California Senate Bill No. 1875. Chapter 816, Statutes of 2000. (Sept. 28, 2000)
14. Cullen DJ, Bates DW, Small SD, Cooper JB, Nemeskal AR, Leape LL. The incident
reporting system does not detect adverse drug events: a problem for quality
improvement. Jt Comm J Qual Improv. 1995; 21: 541-8.
67

15. Bates DW, Boyle DL, Vander Vliet MB, Schneider J, Leape L. Relationship between
medication errors and adverse drug events. J Gen Intern Med. 1995; 10: 199-205.
16. Jha AK, Kuperman GJ, Teich JM, Leape L, Shea B, Rittenberg E, et al. Identifying
adverse drug events: development of a computer-based monitor and comparison with
chart review and stimulated voluntary report. J Am Med Inform Assoc. 1998; 5: 305-14.
17. Adverse drug events: the magnitude of health risk is uncertain because of limited
incidence data. Washington, D.C.: General Accounting Office; 2000.
18. Gandhi TK, Burstin HR, Cook EF, Puopolo AL, Haas JS, Brennan TA, et al. Drug
Complications in Outpatients. J Gen Intern Med. 2000; 15: 149-54.
19. Ash JS, Gorman PN, Hersh WR. Physician order entry in U.S. hospitals. Proceedings
AMIA Annual Symposium. 1998: 235-39.
20. Ringold DJ, Santell JP, Schneider PJ. ASHP national survey of pharmacy practice in
acute care settings: dispensing and administration--1999. Am J Health Syst Pharm. 2000;
57: 1759-75.
21. Overhage JM, Tierney WM, Zhou XH, McDonald CJ. A randomized trial of "corollary
orders" to prevent errors of omission. J Am Med Inform Assoc. 1997; 4: 364-75.
22. Bates DW, Leape LL, Cullen DJ, Laird N, Petersen LA, Teich JM, et al. Effect of
computerized physician order entry and a team intervention on prevention of serious
medication errors . JAMA. 1998; 280: 1311-6.
23. Bates DW, Teich JM, Lee J, Seger D, Kuperman GJ, Ma'Luf N, et al. The impact of
computerized physician order entry on medication error prevention. J Am Med Inform
Assoc. 1999; 6: 313-21.
24. Teich JM, Merchia PR, Schmiz JL, Kuperman GJ, Spurr CD, Bates DW. Effects of
computerized physician order entry on prescribing practices. Arch Intern Med. 2000;
160: 2741-7.
25. Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computer-based clinical decision
support systems on physician performance and patient outcomes: a systematic review.
JAMA. 1998; 280: 1339-46.
26. Walton RT, Harvey E, Dovey S, Freemantle N. Computerised advice on drug dosage to
improve prescribing practice (Cochrane Review). In: The Cochrane Library, Issue 2,
2001. Oxford: Update Software.
27. Evans RS, Classen DC, Pestotnik SL, Lundsgaarde HP, Burke JP. Improving empiric
antibiotic selection using computer decision support. Arch Intern Med. 1994; 154: 878-
84.
68

28. Evans RS, Pestotnik SL, Classen DC, Clemmer TP, Weaver LK, Orme JF, Jr., et al. A
computer-assisted management program for antibiotics and other antiinfective agents. N
Engl J Med. 1998; 338: 232-8.
29. Johnston ME, Langton KB, Haynes RB, Mathieu A. Effects of computer-based clinical
decision support systems on clinician performance and patient outcome. A critical
appraisal of research . Ann Intern Med. 1994; 120: 135-42.
30. Shojania KG, Yokoe D, Platt R, Fiskio J, Ma'luf N, Bates DW. Reducing vancomycin
use utilizing a computer guideline: results of a randomized controlled trial. J Am Med
Inform Assoc. 1998; 5: 554-62.
31. Teich JM, Glaser JP, Beckley RF, Aranow M, Bates DW, Kuperman GJ, et al. Toward
cost-effective, quality care; the Brigham Integrated Computing System. Proc. 2nd
Nicholas E. Davies CPR Recognition Symposium. Washington, D.C.: Computerized
Patient Record Institute. 1996;3-34.; 1996: 3-34.
32. Bates DW, Spell N, Cullen DJ, Burdick E, Laird N, Petersen LA, et al. The costs of
adverse drug events in hospitalized patients. Adverse Drug Events Prevention Study
Group. JAMA. 1997; 277: 307-11.
33. Evans RS, Pestonik SL, Classen DC, Burke JP. Evaluation of a computer-assisted
antibiotic-dose monitor. Ann Pharmacother. 1999; 33: 1026-31.
34. Massaro TA. Introducing physician order entry at a major academic medical center: I.
Impact on organizational culture and behavior. Acad Med. 1993; 68: 20-5.
69

Chapter 7. The Clinical Pharmacist’s Role in Preventing Adverse Drug
Events
Rainu Kaushal, MD, MPH
David W. Bates, MD, MSc
Harvard Medical School
Background
A large literature documents the multiple roles clinical pharmacists can play in a variety
of health care settings. 1-8 Much of this literature focuses on measures of impact not directly
relevant to this Report – eg, economic benefits, 4,8 patient compliance, 6 and drug monitoring. 2,3
More recently, systems-based analyses of medication errors and adverse drug events (ADEs)
have drawn attention to the impact clinical pharmacists can exert on the quality and safety of
medication use. 9-12 In this chapter, we review the evidence supporting the premise that direct
participation of pharmacists’ in clinical care reduces medication errors and ADEs in hospitalized
and ambulatory patients.
Practice Description
Clinical pharmacists may participate in all stages of the medication use process,
including drug ordering, transcribing, dispensing, administering, and monitoring. The specific
activities of clinical pharmacists vary substantially in the studies reviewed in this chapter. In the
hospital setting, one study evaluated the role of a senior pharmacist participating fully in
intensive care unit rounds and available throughout the day in person or by page for questions. 13
Another study evaluated a ward pharmacy service that examined order sheets for new therapies
and carried out checks that were formerly performed in the pharmacy. 14
Pharmacists may also play a role at the time of discharge. One study reported the impact
of clinical pharmacists’ consultation for geriatric patients at the time of discharge, 15 with
pharmacists serving as consultants to physicians and reinforcing patients’ knowledge of their
medication regimen. The roles of clinical pharmacists are similarly diverse in studies of
ambulatory settings. Here they include the provision of consultative services, 16-18 patient
education, 16-18 therapeutic drug monitoring, 3 and even follow-up telephone calls to patients. 3,16-18
Prevalence and Severity of the Target Safety Problem
It is estimated that over 770,000 people are injured or die in hospitals from adverse drug
events (ADEs) annually. 19-21 The few hospitals that have studied incidence rates of ADEs have
documented rates ranging from 2 to 7 per 100 admissions. 11,19,22,23 A precise national estimate is
difficult to calculate due to the variety of criteria and definitions used by researchers. 24 One
study of preventable inpatient ADEs in adults demonstrated that 56% occurred at the stage of
ordering, 34% at administration, 6% at transcribing, and 4% at dispensing. 22 In this study, the
drug class most commonly associated with preventable ADEs was analgesics, followed by
sedatives and antibiotics. Even fewer studies have been conducted in the outpatient setting. One
recent cross-sectional chart review and patient care survey found an ADE rate of 3% in adult
primary care outpatients. 25
Opportunities for Impact
Although many hospital pharmacy departments offer clinical pharmacy consultation
services, 26 the degree to which these services include rounding with clinicians 13 is unclear.
71

Hospital pharmacies provide support to a variable degree in different ambulatory settings
(clinics, nursing homes, adult daycare), 26 but the precise nature of clinical pharmacists’ activities
in these settings is not uniformly characterized.
Study Designs
We identified 3 systematic reviews of clinical pharmacists in the outpatient setting. 5,17,18
We included only the most recent review, 18 as it followed the most rigorous methodology and
included the bibliographies of the previous reviews. 5,17 We identified only one study 16 of the
impact of clinical pharmacists on patient outcomes in the ambulatory setting that had not been
included in this systematic review. 18 This study, a randomized controlled trial evaluating clinical
pharmacists’ participation in the management of outpatients with congestive heart failure, 16 was
therefore also included.
For studies of clinical pharmacists in the hospital setting, an older review 2 did not meet
the characteristics of a systematic review, but was a very thorough summary of the relevant
literature. It included 8 studies evaluating pharmacists’ roles in detecting and reporting ADEs in
the hospital setting. Preliminary review of these studies revealed that the measured outcomes
were Level 3 (detection of ADEs as an end in itself). Consequently, we did not include them.
One older retrospective before-after analysis (Level 3) 14 did meet our inclusion criteria, as did 2
more recent studies of clinical pharmacists’ roles in the inpatient setting: a prospective,
controlled before-after study (Level 2) 13 and a randomized controlled trial (Level 1). 15
We included an additional meta-analysis focusing on therapeutic drug monitoring by
clinical pharmacists in the hospital and ambulatory settings. 3 The studies included in this metaanalysis
consisted predominantly of controlled observational studies (Level 3) and nonrandomized
clinical trials (Level 2), but one randomized controlled trial was also included (Level
1-3A).
Table 7.1 lists the studies reviewed in this chapter and briefly describes their salient
features. All of the listed studies involved adult patients, as the single pediatric study 9 identified
by our literature search had no control group (Level 4 design) and was therefore excluded.
Study Outcomes
Level 1 outcomes reported in the included studies consisted of ADEs 13 and clinical
events related to heart failure, including mortality. 16 One study used telephone interviews to
solicit patient self-reports of adverse drug reactions. 13,27 This study was excluded since the
systematic review of clinical pharmacists’ roles in ambulatory settings 18 incorporated its findings
in several of its analyses.
The distinction between Level 2 and 3 outcomes can be somewhat ambiguous in studies
of prescribing practice, as the exact point at which choices of therapeutic agents or dosing
patterns become not just sub-optimal but actually represent “errors” is difficult to define
precisely. Even for objective outcomes, such as serum drug concentrations, the connection to
patient outcomes is weak in some cases (eg, monitoring vancomycin levels 28,29 ), and therefore
more appropriately designated as Level 3 rather than Level 2. Acknowledging the subjectivity of
this judgment in some cases, we included studies that contained a mixture of Level 2 and 3
outcomes, including “prescribing problems,” (which included inappropriate choice of therapy,
dosage errors, frequency errors, drug-drug interactions, therapeutic duplications, and allergies 15 ),
and serum drug concentrations for a broad range of medications. 3
72

Evidence for Effectiveness of the Practice
In the inpatient setting, Leape’s study 13 demonstrated a statistically significant 66%
decrease in preventable ADEs due to medication ordering. The study of geriatric patients at the
time of discharge demonstrated clinically and statistically significant decreases in medication
errors. 15 The meta-analysis of the effect of clinical pharmacokinetics services on maintaining
acceptable drug ranges indicated only modest effect sizes for the outcomes measured, and only 2
of the main results achieved statistical significance. 3
The comprehensive review of clinical pharmacist services in ambulatory settings reported
positive impacts for patients with hypertension, hypercholesterolemia, chronic heart failure, and
diabetes. 18 However, the authors identify important limitations: these studies are not easily
generalizable, only 2 studies compared pharmacist services with other health professional
services, and both studies had important biases. Consequently, they emphasized the need for
more rigorous research to document the effects of outpatient pharmacist interventions. 18 The
additional study of outpatients demonstrated significant decreases in mortality and heart failure
events, 16 but these results may reflect closer follow-up and monitoring (including telemetry) for
the intervention group or the higher doses of angiotensin-converting enzyme (ACE) inhibitors
the patients received. Generalizing this benefit to other conditions is difficult since most
conditions do not have a single medication-related process of care that delivers the marked
clinical benefits as do ACE inhibitors in the treatment of congestive heart failure. 30
Potential for Harm
Introducing clinical pharmacists might potentially disrupt routine patient care activities.
However, the 2 studies that assessed physician reactions to clinical pharmacists 13,27 found
excellent receptivity and subsequent changes in prescribing behavior.
Costs and Implementation
Two studies examined resource utilization and cost savings in the inpatient setting. The
intensive care unit study indicated that there could be potential savings of $270,000 per year for
this hospital if the intervention involved re-allocation of existing pharmacists’ time and
resources. 13 McMullin et al studied all interventions made by 6 hospital pharmacists over a onemonth
period at a large university hospital and estimated annual cost savings of $394,000. 31
A systematic review of outpatient pharmacists indicated that pharmacist interventions
could lead to increased scheduled service utilization and decreased non-scheduled service
utilization, specialty visits, and numbers and costs of drugs. 18
73

Comment
At present, one study provides strong evidence for the benefit of clinical pharmacists in
reducing ADEs in hospitalized intensive care unit patients. 13 One additional study provides
modest support for the impact of ward-based clinical pharmacists on the safety and quality of
inpatient medication use. 14 The evidence in the outpatient setting is less substantial, and not yet
convincing. Given the other well-documented benefits of clinical pharmacists and the promising
results in the inpatient setting, more focused research documenting the impact of clinical
pharmacist interventions on medication errors and ADEs is warranted.
74

Table 7.1. Studies of clinical pharmacists’ impact on ADEs and medication errors*
Study Study Design Study Outcomes Results
Beney, 2000. 18 Systematic
review of the roles and impacts
of pharmacists in ambulatory
settings; reviewed studies
included 16,000 outpatients
and 40 pharmacists
Gattis, 1999. 16 181 patients
with heart failure due to left
ventricular dysfunction
followed in a general
cardiology clinic
Leape, 1999. 13 Medical and
cardiac intensive care unit
patients at Massachusetts
General Hospital, a tertiary
care hospital in Boston
Leach, 1981. 14 315 patients at
Queen Elizabeth Hospital in
Birmingham, England
Lipton, 1992. 15 236 geriatric
patients discharged from the
hospital on three or more
medications
Ried, 1989. 3 Pooled patient
population not reported, but
review of articles indicates
predominance of studies of
(mostly adult) hospitalized
patients
Level 1A
(systematic
review)
Level 1
(RCT)
Level 2
(prospective
before-after study
with concurrent
control)
Level 3
(retrospective
before-after
analysis)
Levels 1-3
(variety of patient
outcomes, surrogate
outcomes, impacts on
physician prescribing
practices and measures of
resource use)
Level 1
(mortality and other
clinical outcomes related
to heart failure)
75
Improvement in outcomes for
patients with hypertension,
hypercholesterolemia,
chronic heart failure, and
diabetes
16 versus 4 deaths or other
heart failure events (p

References
1. Dyer CC, Oles KS, Davis SW. The role of the pharmacist in a geriatric nursing home: a
literature review. Drug Intell Clin Pharm. 1984;18:428-433.
2. Hatoum HT, Catizone C, Hutchinson RA, Purohit A. An eleven-year review of the
pharmacy literature: documentation of the value and acceptance of clinical pharmacy. Drug
Intell Clin Pharm. 1986;20:33-48.
3. Ried LD, McKenna DA, Horn JR. Meta-analysis of research on the effect of clinical
pharmacokinetics services on therapeutic drug monitoring. Am J Hosp Pharm.
1989;46:945-951.
4. Willett MS, Bertch KE, Rich DS, Ereshefsky L. Prospectus on the economic value of
clinical pharmacy services. A position statement of the American College of Clinical
Pharmacy. Pharmacotherapy. 1989;9:45-56.
5. Carter BL, Helling DK. Ambulatory care pharmacy services: the incomplete agenda. Ann
Pharmacother. 1992;26:701-708.
6. Tett SE, Higgins GM, Armour CL. Impact of pharmacist interventions on medication
management by the elderly: a review of the literature. Ann Pharmacother. 1993;27:80-86.
7. Jenkins MH, Bond CA. The impact of clinical pharmacists on psychiatric patients.
Pharmacotherapy. 1996;16:708-714.
8. Schumock GT, Meek PD, Ploetz PA, Vermeulen LC. Economic evaluations of clinical
pharmacy services–1988-1995. The Publications Committee of the American College of
Clinical Pharmacy. Pharmacotherapy. 1996;16:1188-1208.
9. Folli HL, Poole RL, Benitz WE, Russo JC. Medication error prevention by clinical
pharmacists in two children's hospitals. Pediatrics. 1987;79:718-722.
10. Lesar TS, Briceland LL, Delcoure K, Parmalee JC, Masta-Gornic V, Pohl H. Medication
prescribing errors in a teaching hospital. JAMA. 1990;263:2329-2334.
11. Bates DW, Boyle DL, Vander Vliet MB, Schneider J, Leape L. Relationship between
medication errors and adverse drug events. J Gen Intern Med. 1995;10:199-205.
12. Kaushal R, Bates DW, Landrigan C, McKenna J, Clapp MD, Federico F, et al. medication
errors and adverse drug events in pediatric inpatients. JAMA. 2001;285:2114-2120.
13. Leape LL, Cullen DJ, Clapp MD, Burdick E, Demonaco HJ, Erickson JI, et al. Pharmacist
participation on physician rounds and adverse drug events in the intensive care unit. JAMA.
1999;282:267-270.
14. Leach RH, Feetam C, Butler D. An evaluation of a ward pharmacy service. J Clin Hosp
Pharm. 1981;6:173-182.
15. Lipton HL, Bero LA, Bird JA, McPhee SJ. The impact of clinical pharmacists'
consultations on physicians' geriatric drug prescribing. A randomized controlled trial. Med
Care. 1992;30:646-658.
16. Gattis WA, Hasselblad V, Whellan DJ, O'Connor CM. Reduction in heart failure events by
the addition of a clinical pharmacist to the heart failure management team: results of the
Pharmacist in Heart Failure Assessment Recommendation and Monitoring (PHARM)
Study. Arch Intern Med. 1999;159:1939-1945.
17. Hatoum HT, Akhras K. 1993 Bibliography: a 32-year literature review on the value and
acceptance of ambulatory care provided by pharmacists. Ann Pharmacother.
1993;27:1106-1119.
76

18. Beney J, Bero LA, Bond C. Expanding the roles of outpatient pharmacists: effects on
health services utilisation, costs, and patient outcomes. In: The Cochrane Library, Issue 2,
2000. Oxford: Update Software.
19. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in
hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA.
1997;277:301-306.
20. Cullen DJ, Bates DW, Small SD, Cooper JB, Nemeskal AR, Leape LL. The incident
reporting system does not detect adverse drug events: a problem for quality improvement.
Jt Comm J Qual Improv. 1995;21:541-548.
21. Cullen DJ, Sweitzer BJ, Bates DW, Burdick E, Edmondson A, Leape LL. Preventable
adverse drug events in hospitalized patients: a comparative study of intensive care and
general care units. Crit Care Med. 1997;25:1289-1297.
22. Bates DW, Cullen DJ, Laird N, Petersen LA, Small SD, Servi D, et al. Incidence of adverse
drug events and potential adverse drug events: Implications for prevention. JAMA.
1995;274:29-34.
23. Jha AK, Kuperman GJ, Teich JM, Leape L, Shea B, Rittenberg E, et al. Identifying adverse
drug events: development of a computer-based monitor and comparison with chart review
and stimulated voluntary report. J Am Med Inform Assoc. 1998;5:305-314.
24. Adverse drug events: the magnitude of health risk is uncertain because of limited incidence
data. Washington, DC: General Accounting Office; 2000.
25. Gandhi TK, Burstin HR, Cook EF, Puopolo AL, Haas JS, Brennan TA, et al. Drug
complications in outpatients. J Gen Intern Med. 2000;15:149-154.
26. Ringold DJ, Santell JP, Schneider PJ. ASHP national survey of pharmacy practice in acute
care settings: dispensing and administration–1999. Am J Health Syst Pharm.
2000;57:1759-1775.
27. Hanlon JT, Weinberger M, Samsa GP, Schmader KE, Uttech KM, Lewis IK, et al. A
randomized, controlled trial of a clinical pharmacist intervention to improve inappropriate
prescribing in elderly outpatients with polypharmacy. Am J Med. 1996;100:428-437.
28. Freeman CD, Quintiliani R, Nightingale CH. Vancomycin therapeutic drug monitoring: is
it necessary? Ann Pharmacother. 1993;27:594-598.
29. Cantu TG, Yamanaka-Yuen NA, Lietman PS. Serum vancomycin concentrations:
reappraisal of their clinical value. Clin Infect Dis. 1994;18:533-543.
30. Flather MD, Yusuf S, Kober L, Pfeffer M, Hall A, Murray G, et al. Long-term ACEinhibitor
therapy in patients with heart failure or left-ventricular dysfunction: a systematic
overview of data from individual patients. ACE-Inhibitor Myocardial Infarction
Collaborative Group. Lancet. 2000;355:1575-1581.
31. McMullin ST, Hennenfent JA, Ritchie DJ, Huey WY, Lonergan TP, Schaiff RA, et al. A
prospective, randomized trial to assess the cost impact of pharmacist-initiated
interventions. Arch Intern Med. 1999;159:2306-2309.
77

Chapter 8. Computer Adverse Drug Event (ADE) Detection and Alerts
Tejal K. Gandhi, MD, MPH
David W. Bates, MD, MSc
Harvard Medical School
Background
Adverse drug events (ADEs) occur in both inpatient and outpatient settings. 1,2 Most
institutions use spontaneous incident reporting to detect adverse events in general, and ADEs in
particular. Spontaneous incident reporting relies exclusively on voluntary reports from nurses,
pharmacists and physicians focused on direct patient care. However, spontaneous reporting is
ineffective, identifying only one in 20 ADEs. 3 Efforts to increase the frequency of spontaneous
reporting have had only a minor impact.
Several studies demonstrate the effectiveness of using computerized detection and alert
systems (referred to as computer “monitors”) to detect ADEs. In 1991, Classen et al 4 published
information about a computerized ADE monitor that was programmed to identify signals—in
effect mismatches of clinical information—that suggested the presence of an ADE. The signals
included sudden medication stop orders, antidote ordering, and certain abnormal laboratory
values. 4 The computerized signals were then evaluated by a pharmacist who determined whether
an ADE had occured. Based on the rules of this study, Jha et al developed a similar monitor that
identified approximately half the ADEs identified by chart review, at much lower cost. 5
Similarly, Bowman and Carlstedt used the Regenstrief Medical Record System to create a
computerized inpatient ADE detection system. 6 Compared to the “gold standard” of chart
review, the monitor had 66% sensitivity and 61% specificity, with a positive predictive value of
0.34. Finally, one community hospital implemented an ADE monitor with triggers that were
reviewed by pharmacists who then contacted physicians to make appropriate regimen changes.
This study identified opportunities to prevent patient injury at a rate of 64/1000 admissions. 7
These studies and others demonstrate the potential value of computer monitors, especially when
linked to effective integrated information systems. While monitors are not yet widely used, they
offer an efficient approach for monitoring the frequency of ADEs on an ongoing basis, and the
Health Care Financing Administration is considering mandating them. 8
Practice Description
Computerized ADE alert monitors use rule sets to search signals that suggest the
presence of adverse drug events. The most frequently studied rule sets (or “triggers”) are those
that search for drug names (eg, naloxone, kayexalate), drug-lab interactions (eg, heparin and
elevated PTT) or lab levels alone (eg, elevated digoxin levels) that frequently reflect an ADE.
Simple versions can be implemented with pharmacy and laboratory data alone, although the
yield and positive predictive value of signals is higher when the 2 databases are linked.
Further refinements include searches for International Classification of Diseases (ICD-9)
codes, and text-searching of electronic nursing bedside charting notes or outpatient notes for
drug-symptom combinations (eg, medication list includes an angiotensin converting enzyme
inhibitor and the patient notes mention “cough”). Although these refinements do increase the
yield of monitors, they require linkage to administrative databases or electronic medical records.
The information captured with computer monitors is used to alert a responsible clinician
or pharmacist, who can then change therapy based on the issue in question. Systems are designed
79

to alert the monitoring clinician in various ways. Alerts can go to one central location (eg,
hospital pharmacist) or to individual physicians. Monitoring pharmacists typically review the
alert and contact the appropriate physician if they determine that the alert has identified a true
event. The alert modality also varies based on the available technology, from printed out reports,
to automatic paging of covering physicians, to display of alerts on computer systems (either in
results or ordering applications). It should be emphasized that computerized physician order
entry (Chapter 6) is not a requirement for these monitors. Thus, a simple version of this approach
could be implemented in most US hospitals.
Prevalence and Severity of the Target Safety Problem
It is estimated that over 770,000 people are injured or die in hospitals from ADEs
annually, 3,9,10 but variations in study criteria and definitions prevent precise national estimates. 11
Fewer data address the epidemiology of ADEs in the outpatient setting. One recent study found
an ADE rate of 3% in adult primary care outpatients, 12 while an older study reported a similar
rate of 5% among ambulatory patients attending an internal medicine clinic over a 1-year
period. 2
Detection and alerting interventions primarily target ADEs related to the medication
ordering process. One study of preventable inpatient ADEs in adults demonstrated that 56%
occurred at the stage of ordering. 13 Among the 6.5 ADEs per 100 admissions in this study, 28%
were judged preventable, principally by changing the systems by which drugs are ordered and
administered. 14 In one study of computerized ADE detection, the ADEs identified by
computerized monitoring were significantly more likely to be classified as severe than those
identified by chart review (51% vs. 42%, p=0.04). 5 Thus, monitors may capture a subset of
events with the most potential for patient injury.
Injuries due to drugs have important economic consequences. Inpatients that suffer ADEs
have increased lengths of stay of nearly 2 days and added hospital costs of more than $2000. 9,15
Bootman has estimated the annual cost of drug-related morbidity and mortality within the United
States at $76.6 billion, with the majority ($47 billion) related to hospital admissions due to drug
therapy or the absence of appropriate drug therapy. 16
Opportunities for Impact
Unfortunately, there are no good data as to how many hospitals have integrated lab and
medication systems, which are required for many of the triggers used in computerized ADE
monitors.
Study Designs
We found 5 studies of computerized ADE alert systems that were Level 3 or better (see
Table 8.1). Four studies 17-20 detected potential ADEs using computerized monitoring and then
alerted physicians or pharmacists about the event. One additional study 21 both alerted the
monitoring clinician and made recommendations for actions relating to the suspect condition.
Four studies 17-20 were in the inpatient setting and one was in the ambulatory setting. 21
Study Outcomes
All of the studies reported Level 1 or 2 outcomes. Level 1 outcomes were the rate of
ADEs 18,19 and renal impairment (as reflected by rises in creatinine). 20 Level 2 outcomes included
percent of time recommended actions were taken and time to respond to an event. 17-20
80

Evidence for Effectiveness of the Practice
One study demonstrated significant decreases in adverse clinical outcomes with the alert
systems. 18 This decrease included a surprisingly large reduction in allergic reactions (ones not
previously known); it is not clear how the computer alert or decision support system could have
contributed to this result. 18 The second study 19 showed no significant difference in the number of
adverse events in the intervention and control groups. This study and 3 others revealed
significant improvements in response times concerning lab values, 17-20 and one study found a
significant decrease in the risk of serious renal impairment. 20 Finally, one study demonstrated
significant changes in physician behavior/modification of therapy based on alerts with
recommended actions. 21 The effect sizes shown in these studies are listed in Table 8.1.
Potential for Harm
None of the studies discuss any potential for harm associated with the monitor and alerts.
It is certainly possible that alerts could be erroneous, but it is doubtful that this would lead to any
direct patient harm. As in studies of safety in other industries, one possible source of harm could
be too many false positives. Large numbers of clinically insignificant warnings for patients
would interfere with routine care, and might result in providers ignoring all warnings, even
clinically meaningful ones.
Costs and Implementation
In general, implementation of alert monitors requires computer systems that can link
laboratory and medication information. Integrating this information requires the creation of
interface between the drug and laboratory databases, with costs that will vary depending on the
nature of the existing information systems. In addition, alert methods vary, with some
applications directly contracting physicians (which requires further integration of coverage and
pager databases) and others using pharmacist intermediaries. The cost of pharmacist review of
triggers was less than 1 FTE per hospital in 2 studies 5,7 ; one of them reported an annual cost
savings of up to 3 million dollars by reducing preventable ADEs with this alerting technique. 7
Studies thus far suggest that physicians view computerized alert systems favorably.
Forty-four percent of physician-respondents receiving alerts indicated that the alerts were helpful
and 65% wished to continue receiving them (although these alerts went to many physicians
because it was unclear who the responsible doctor was). In another study in which alerts were
sent only to the responsible physician, 95% of physician-respondents were pleased to receive
them. 19
The systems in these studies were all “homegrown” and contained idiosyncrasies that
might undermine their implementation elsewhere. Clearly it is important that systems track
which physicians are responsible for which patients. In addition, all 4 of the inpatient studies
were conducted at tertiary care hospitals and the outpatient study was done at clinics affiliated
with a tertiary care center. The translation of this alerting approach to community settings may
be difficult. One community teaching hospital has reported success in detecting opportunities to
prevent injury (64/1000 admissions) using computerized detection and alerting. This report had
only a Level 4 design (no control group), so it was not included in the Table. 7
81

Comment
Computerized real-time monitoring facilitates detection of actual and potential ADEs and
notification of clinicians. This in turn may aid in the prevention of ADEs or decrease the chances
that ADEs will cause harm. The monitors also yield improvements in secondary measures
relating to the length of time until response and the quality of response.
The applications in these studies were all “homegrown.” Future applications should be
evaluated and refined further. In particular, it is important to quantify the yield of collecting
these data in terms of patient outcomes, since the start-up costs are significant. If monitors do
lead to important clinical benefits, they should become standard features of commercially
available hospital computer systems. As this occurs, careful attention will need to be paid to
optimizing the response process.
In addition, little has been done to translate these monitoring systems to the outpatient
setting, largely because outpatient clinical information is often not computerized or resides in
disparate systems. As integrated computerized outpatient records become more common, the
systems developed in the inpatient setting should be translatable to the outpatient setting.
82

Table 8.1. Included studies of computerized systems for ADE detection and alerts*
Study Intervention Study Design,
Outcomes
Kuperman, 1999 19
Computerized alerts to physicians
via paging system
McDonald, 1976 21
Alerts to outpatient physicians
with suggested responses to
medication related events
Evans, 1994 18
Computerized monitor to detect
ADEs (including drug/lab
monitors and searches of nursing
notes) and then computerized
alerts to physicians
Rind, 1994 20
Computerized alert system to
physicians about rising serum
creatinine values in patients
receiving potentially nephrotoxic
medications
Bradshaw, 1989 17
Computerized alerts integrated
into result review and alerts by
flashing light
Level 1,
Level 1
Level 1,
Level 2
Level 3,
Level 1
Level 3,
Level 1
Level 3,
Level 2
83
Results†
Median time until initial treatment ordered:
1 vs. 1.6 hours (p=0.003)
Median time until condition resolved:
8.4 vs. 8.9 hours (p=0.11)
Number of adverse events:
no significant difference
Physicians performed recommended testing:
36% vs. 11% (p

References
1. Kellaway GS, McCrae E. Intensive monitoring for adverse drug effects in patients
discharged from acute medical wards. N Z Med J. 1973;78:525-528.
2. Hutchinson TA, Flegel KM, Kramer MS, Leduc DG, Kong HH. Frequency, severity and
risk factors for adverse drug reactions in adult out-patients: a prospective study. J Chronic
Dis. 1986;39:533-542.
3. Cullen DJ, Bates DW, Small SD, Cooper JB, Nemeskal AR, Leape LL. The incident
reporting system does not detect adverse drug events: a problem for quality improvement.
Jt Comm J Qual Improv. 1995;21:541-548.
4. Classen DC, Pestotnik SL, Evans RS, Burke JP. Computerized surveillance of adverse drug
events in hospital patients. JAMA. 1991;266:2847-2851.
5. Jha AK, Kuperman GJ, Teich JM, Leape L, Shea B, Rittenberg E, et al. Identifying adverse
drug events: development of a computer-based monitor and comparison with chart review
and stimulated voluntary report. J Am Med Inform Assoc. 1998;5:305-314.
6. Bowman L, Carlstedt BC, Black CD. Incidence of adverse drug reactions in adult medical
inpatients. Can J Hosp Pharm. 1994;47:209-216.
7. Raschke RA, Gollihare B, Wunderlich TA, Guidry JR, Leibowitz AI, Peirce JC, et al. A
computer alert system to prevent injury from adverse drug events: development and
evaluation in a community teaching hospital. JAMA. 1998;280:1317-20. Published erratum
appears in JAMA 1999 Feb 3;281:420.
8. Department of Health and Human Services. Health care financing administration. Fed
Regist. 1997; 62.
9. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in
hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA.
1997;277:301-306.
10. Cullen DJ, Sweitzer BJ, Bates DW, Burdick E, Edmondson A, Leape LL. Preventable
adverse drug events in hospitalized patients: a comparative study of intensive care and
general care units. Crit Care Med. 1997;25:1289-1297.
11. Adverse drug events: the magnitude of health risk is uncertain because of limited incidence
data. Washington, DC: General Accounting Office; 2000. GAO/HEHS-00-21.
12. Gandhi TK, Burstin HR, Cook EF, Puopolo AL, Haas JS, Brennan TA, et al. Drug
complications in outpatients. J Gen Intern Med. 2000;15:149-154.
13. Bates DW, Cullen DJ, Laird N, Petersen LA, Small SD, Servi D, et al. Incidence of adverse
drug events and potential adverse drug events. Implications for prevention. ADE
Prevention Study Group. JAMA. 1995;274:29-34.
14. Leape LL, Bates DW, Cullen DJ, Cooper J, Demonaco HJ, Gallivan T, et al. Systems
analysis of adverse drug events. ADE Prevention Study Group. JAMA. 1995;274:35-43.
15. Bates DW, Spell N, Cullen DJ, Burdick E, Laird N, Petersen LA, et al. The costs of
adverse drug events in hospitalized patients. Adverse Drug Events Prevention Study
Group. JAMA. 1997;277:307-311.
16. Johnson JA, Bootman JL. Drug-related morbidity and mortality. A cost-of-illness model.
Arch Intern Med. 1995;155:1949-1956.
17. Bradshaw KE, Gardner RM, Pryor TA. Development of a computerized laboratory alerting
system. Comput Biomed Res. 1989;22:575-587.
18. Evans RS, Pestotnik SL, Classen DC, Horn SD, Bass SB, Burke JP. Preventing adverse
drug events in hospitalized patients. Ann Pharmacother. 1994;28:523-527.
84

19. Kuperman GJ, Teich JM, Tanasijevic MJ, Ma'Luf N, Rittenberg E, Jha A, et al. Improving
response to critical laboratory results with automation: results of a randomized controlled
trial. J Am Med Inform Assoc. 1999;6:512-522.
20. Rind DM, Safran C, Phillips RS, Wang Q, Calkins DR, Delbanco TL, et al. Effect of
computer-based alerts on the treatment and outcomes of hospitalized patients. Arch Intern
Med. 1994;154:1511-1517.
21. McDonald CJ. Use of a computer to detect and respond to clinical events: its effect on
clinician behavior. Ann Intern Med. 1976;84:162-167.
85

Chapter 9. Protocols for High-Risk Drugs: Reducing Adverse Drug Events
Related to Anticoagulants
Tejal K. Gandhi, MD, MPH
Harvard Medical School
Kaveh G. Shojania, MD
University of California, San Francisco School of Medicine
David W. Bates, MD, MSc
Harvard Medical School
Background
Published studies of adverse drug events and multiple case reports have consistently
identified certain classes of medications as particularly serious threats to patient safety. 1-3 These
“high risk” medications include concentrated electrolyte solutions such as potassium chloride,
intravenous insulin, chemotherapeutic agents, intravenous opiate analgesics, and anticoagulants
such as heparin and warfarin. Analyses of some of the adverse events involving these mediations
have led to important recommendations regarding their administration. Examples include the use
of order templates for chemotherapeutic agents, removal of intravenous electrolyte solutions
from general ward stock, and protocols for reviewing the settings of intravenous pumps
delivering continuous or frequent doses of opiates. 2,4,5 While these recommendations have high
face validity, they have generally not been subject to formal evaluation regarding their impact in
reducing the targeted adverse events. By contrast, several practices relating to the management
of patients receiving anticoagulants have been evaluated quite extensively, and therefore
constitute the focus of this chapter.
Heparin and warfarin are medications whose use or misuse carry significant potential for
injury. Subtherapeutic levels can lead to thromboembolic complications in patients with atrial
fibrillation or deep venous thrombosis (DVT), while supratherapeutic levels can lead to bleeding
complications. These medications are commonly involved in ADEs for a variety of reasons,
including the complexity of dosing and monitoring, patient compliance, numerous drug
interactions, and dietary interactions that can affect drug levels. Strategies to improve both the
dosing and monitoring of these high-risk drugs have potential to reduce the associated risks of
bleeding or thromboembolic events.
Practice Description
The practices reviewed in this chapter are all intended to reduce dosing and/or
monitoring errors for heparin and warfarin, as follows:
• Heparin dosing protocols (“nomograms”) typically involve a standard initial
bolus and infusion rate, instructions for when to draw the first partial
thromboplastin time (PTT), and orders for dosing adjustments in response to
this and subsequent values (so nurses can adjust doses automatically). In some
cases, the initial bolus and infusion rates are based on patient weight.
• Inpatient anticoagulation services for both heparin and warfarin (with or
without dosing nomograms) typically consist of pharmacist-run services that
provide daily pharmacy input on dosing and monitoring for patients on heparin
and/or warfarin. (We excluded studies focusing solely on warfarin prophylaxis
in orthopedic patients. 6 )
87

• Outpatient anticoagulation clinics provide coordinated services for managing
outpatient warfarin therapy. Services typically include anticoagulation
monitoring and follow-up, warfarin dose adjustment, and patient education.
These clinics are usually run by pharmacists or nurses operating with physician
back-up, and sometimes following specific dosing nomograms.
• Patient self-monitoring using a home finger-stick device and self-adjustment of
warfarin dosages using a nomogram. (The accuracy of these devices and the
comparability of patients’ and professional readings have been extensively
evaluated. 7-11 )
Prevalence and Severity of the Target Safety Problem
Intravenous heparin and oral warfarin are commonly used medications for cardiac disease
and thromboembolism in the inpatient and outpatient settings. While in the aggregate they are
highly beneficial (see Chapter 31), these drugs can have significant morbidities unless they are
dosed and monitored appropriately. For example, inadequate therapeutic dosing of heparin can
lead to increased length of stay and the potential for clot formation and/or propagation. 12 The
risk of recurrent thromboembolism is reduced if the therapeutic effect of heparin is achieved
quickly. 12 In addition, Landefeld et al 13 showed that the frequency of fatal, major, and minor
bleeding during heparin therapy was twice that expected without heparin therapy. The effect
with warfarin therapy was even more pronounced - approximately 5 times that expected without
warfarin therapy. Consistent with this finding, anticoagulants accounted for 4% of preventable
ADEs and 10% of potential ADEs in one large inpatient study. 1 Finally, careful drug monitoring
in hospitals can reduce ADEs, suggesting that some events are due to inadequate monitoring of
therapies and doses. 14 These studies highlight the clear need for safety-related interventions with
respect to both the dosing and monitoring of these high-risk drugs in order to prevent
thromboembolic and bleeding complications.
Opportunities for Impact
The number of hospitals using weight-based heparin nomograms, or that have established
anticoagulation clinics or services is unknown. Although common in some European countries, 15
patient self-management of long-term anticoagulation with warfarin is unusual in the United
States as many payers, including Medicare, do not currently cover the home testing technology. 15
Study Designs
Heparin nomograms were evaluated in one randomized controlled trial (Level 1), 16 one
prospective cohort comparison (Level 2) 17 and 4 controlled observational studies (Level 3). 18-21
Two of these studies involved weight-based nomograms. 16,21 A third study involving a weightbased
nomogram 22 was included with the studies of anticoagulation services (see below), as
clinical pharmacists actively managed the dosing protocol. We excluded one retrospective
before-after analysis of a weight-based heparin protocol for cardiac intensive care patients, 23
because the method of selecting charts for review was never stated. Moreover, when the authors
found an increase in the number of patients with excessive anticoagulation in the intervention
group, they chose a second group of control patients (again with an unspecified selection
method) for review, and in the end concluded that the difference was not significant.
All studies of outpatient anticoagulation clinics have been Level 3 studies, typically
retrospective before-after analyses, 22,24-28 although one study might more appropriately be
88

egarded as a case-control study. 29 A comprehensive review of the literature on various forms of
anticoagulation management 30 did not meet the criteria for a systematic review, but referenced
all of the additional studies of anticoagulation clinics that we could identify 31-36 and used
quantitative methods to pool their results. We use the pooled results from this article 30 in Table
9.2 in place of individual entries for each of these six Level 3 studies.
Two studies evaluated the impact of a coordinated inpatient anticoagulation service
(with or without nomograms for dosing). 22,37
Patient self-management of warfarin therapy has been evaluated in at least 3 randomized
controlled trials 38-40 (Level 1) and one non-randomized clinical trial. 41 Because a number of
higher-level studies exist, we did not include retrospective cohort analyses (Level 3) addressing
this topic. 42-45
Study Outcomes
Most studies did not evaluate bleeding complications or had insufficient numbers of
patients to evaluate this outcome adequately. One recent study of an anticoagulation clinic’s
adverse events 25 focused on anticoagulation as the primary outcome (Level 1), as did the review
that pooled results from 6 observational studies of anticoagulation clinics. 30 As shown in Tables
9.1-3, the rest of the studies reported Level 2 outcomes, consisting of various indicators of time
to therapeutic anticoagulation and intensity or appropriateness of anticoagulation.
Evidence for Effectiveness of the Practice
• Heparin nomograms: As shown in Table 9.1, all studies showed a significant
decrease (ie, improvement) in time to achievement of a therapeutic PTT and/or
an increase in the proportion of patients in the therapeutic range.
• Inpatient anticoagulation services: As shown in Table 9.2, both Level 3 studies
evaluating this practice showed significant improvements in relevant measures
22, 37
of anticoagulation.
• Outpatient anticoagulation services for warfarin (with and without dosing
nomograms): the multiple Level 3 studies of this practice showed improvements
in relevant measures of anticoagulation, with one exception. 28 This study took
place in a semi-rural region of England, and the hospital-based anticoagulation
clinic was staffed mainly by junior physician trainees rotating through the
clinic. The one study that focused primarily on Level 1 outcomes 25 showed
significant reductions in adverse events related to under- or overanticoagulation.
• Patient self-management: Patient self-management achieved superior measures
of anticoagulation on one Level 1 comparison with routine care. 22,37 More
impressive is that two Level 1 studies 38,46 and one Level 2 study 41 reported
equivalent or superior measures of anticoagulation for self-management
compared with anticoagulation clinics.
Potential for Harm
Heparin nomograms are primarily intended to achieve PTT values within the therapeutic
range as quickly as possible. Although none of the studies showed increased bleeding as a result
89

of aggressive anticoagulation, it is important to note that 4 of the 6 studies showed a significant
increase in the proportion of patients with PTTs above the target range. 16,19-21
Anticoagulation clinics carry the usual theoretical risk that increased fragmentation of
care will introduce new hazards, but no study showed any significant cause for concern.
Patient self-monitoring clearly carries with it risks relating to the possibilities of patient
misunderstanding of, or non-compliance with dosing and monitoring protocols. No increases in
adverse events were observed in the studies reviewed, but the patients evaluated in these studies,
even if randomized, were still chosen from a group of relatively compliant and motivated
patients.
Costs and Implementation
For anticoagulation clinics, one study showed reduced costs of $162,058 per 100 patients
annually, primarily through reductions in warfarin-related hospitalizations and emergency room
visits. 25 Other studies indicate potential cost-savings due to reduced hospitalizations from
anticoagulation-related adverse events, or show that the anticoagulation was revenue
neutral. 19,24,29 Considering without these offsetting potential savings, however, anticoagulant
clinics often require institutional subsidy since professional fee or laboratory payments do not
fully cover costs.
Heparin nomograms may increase lab costs due to more frequent monitoring, but one
study calculated that lab costs were offset by the need for fewer heparin boluses. 22
For patient self-management of warfarin, one study showed that the cost of selfmonitoring
was $11/international normalized ratio (INR) value and postulated that this would be
cost-effective if it reduced the number of clinic visits. 39 Other studies have suggested that the
capillary blood testing devices themselves 47 and the overall practice of patient self-management
are cost-effective. 48,49 In the United States, the home monitoring devices sell for approximately
$1000. Factoring in the price of cartridges and assuming the devices operate without requiring
repair for 5 years, one source estimated an annual cost of approximately $600. 40
Implementation of a heparin nonogram appears feasible, and was well received by
physicians and nurses. 18 Physician/staff education about the protocols was important to its
success. 23,24 One study showed a high level of physician and patient satisfaction with an
anticoagulation clinic. 24 In addition, multiple studies reveal that patients who self-manage
warfarin have significantly higher levels of satisfaction and experience less anxiety. 9,10,38,39
90

Comment
The primary purpose of heparin nomograms is the timely achievement of therapeutic
anticoagulation, and their superiority in this regard (compared with routine care) has been
convincingly established. While none of the studies showed adverse consequences of this focus
on timely anticoagulation, the trend toward increases in excessive anticoagulation presents safety
concerns. Studies powered to detect significant differences in bleeding complications in patients
being managed with heparin dosing protocols may be warranted.
The literature on anticoagulation clinics consists entirely of observational studies with
important possible confounders. Nonetheless, with one exception 28 they are consistently shown
to achieve superior measures of anticoagulation, and in one study, 25 superior clinical outcomes.
Among the practices reviewed in this chapter, the literature on patient self-management is
perhaps the most impressive. Three randomized trials and one non-randomized clinical trial
show that patient control of anticoagulation is at least equivalent, if not superior, to management
by usual care or an anticoagulation clinic. Additional observational studies reach the same
results. 42-45 Thus, a relatively substantial literature supports patient self-management for
outpatient warfarin therapy for motivated patients able to comply with the monitoring and dosing
protocols. These studies clearly involved select groups of patients, 9 so that a larger randomized
trial with intention-to-treat analysis would be helpful.
Many insurance carriers in the United States, including Medicare, do not currently
subsidize the home testing technology or provide only partial coverage. 15 Despite the relatively
high cost of the home testing devices, this practice may nonetheless be cost-effective due to
reduced use of other clinical services. 48,49 A larger US study or detailed cost-effectiveness
analysis appears warranted, especially given the higher level of patient satisfaction with this
approach as compared with outpatient anticoagulation.
91

Table 9.1. Studies focused primarily on heparin or warfarin nomograms*
Study Study Design, Outcomes Results†
Raschke, 1993 16
Weight-based heparin
nomogram for patients
with venous
thromboembolism or
unstable angina
Elliott, 1994 17
Use of heparin
nomogram for patients
with acute proximal
deep venous
thrombosis
Brown, 1997 21
Weight-based heparin
nomogram for ICU
patients requiring
acute anticoagulation
with unfractionated
heparin
Cruickshank, 1991 18
Heparin nomogram for
patients with acute
venous
thromboembolism
Randomized controlled trial
(Level 1)
Various markers of adequate
anticoagulation (Level 2)
Non-randomized clinical trial
(Level 2)
Time to therapeutic PTT
(Level 2)
Retrospective before-after
analysis (Level 3)
Time to therapeutic PTT
(Level 2)
Retrospective before-after
analysis (Level 3)
Time to first therapeutic PTT,
time to correct subsequent
PTTs, time outside the
therapeutic range (Level 2)
92
PTT in therapeutic range within 24
hours: 97% vs. 77% (p

Table 9.1. Studies focused primarily on heparin or warfarin nomograms* (cont.)
Study Study Design, Outcomes Results†
Hollingsworth, 1995 19
Heparin nomogram for
hospitalized patients
with acute venous
thromboembolism
Phillips, 1997 20
Inpatient heparin and
warfarin nomograms
and monitoring charts
Retrospective before-after
analysis (Level 3)
Primary outcome of the study
was length of hospital stay
(Level 3) but time to
therapeutic PTT was a
secondary outcome (Level 2)
Retrospective before-after
analysis (Level 3)
Measures of under- and overanticoagulation
(Level 2)
93
Time to therapeutic PTT: 17.9 vs. 48.8
hours (p

Table 9.2. Inpatient anticoagulation services and outpatient anticoagulation clinics*
Study Study Design, Outcomes Results
Ansell, 1996 30
Pooled comparison of
anticoagulation clinics and
routine medical care
Hamby, 2000 29
Analysis of adverse events
related to outpatient
warfarin therapy among 395
patients followed at a
Veterans Affairs Hospital,
with 306 enrolled in an
anticoagulation clinic and
89 patients receiving usual
care
Lee, 1996 26
Comparison of pharmacistmanaged
anticoagulation
clinic with patient receiving
usual care
Ellis, 1992 37
Pharmacy-managed
inpatient anticoagulation
service (flow sheet for
monitoring, but no
nomogram) for monitoring
patients receiving warfarin
for a variety of indications
Pooled results from 6 Level
3 study designs comparing
anticoagulation clinics with
routine medical care 31-36
(Level 3A)
Major bleeding and
thromboembolic events
(Level 1)
Case-control study
(Level 3)
Adverse events related to
under- or overanticoagulation
(Level 1)
Retrospective cohort
comparison (Level 3)
Hospital admissions related
to under- or overanticoagulation
– ie,
thromboembolic or bleeding
events (Level 1) †
Retrospective before-after
analysis (Level 3)
Anticoagulation “stability”
at discharge and odds of
therapeutic anticoagulation
at first outpatient visit
(Level 2)
94
Major bleeding events per patientyear:
anticoagulation clinic, 0.028
(95% CI: 0-0.069) vs. routine care,
0.109 (95% CI: 0.043-0.268)
Thromboembolic events per
patient-year: anticoagulation
clinic, 0.024 (95% CI: 0-0.08) vs.
routine care, 0.162 (95% CI:
0.062-0.486)
Among the 12 patients with
preventable adverse events related
to anticoagulation, 8 were not
enrolled in the anticoagulation
clinic
Patients receiving usual care had
20 times the relative risk (95% CI:
6-62) of an adverse event
compared with patients in the
anticoagulation clinic.
Patients in anticoagulation clinic
had non-significant reductions in
hospital admissions related to
thromboembolic or bleeding
events compared with control
group ‡
Patients receiving the intervention
were more likely to have PT
“stability” at discharge: 61.5% vs.
42.3% (p=0.02)
Odds of having therapeutic PT at
first outpatient clinic visit with
intervention: OR 5.4 (95% CI:
1.87-15.86)

Table 9.2. Inpatient anticoagulation services and outpatient anticoagulation clinics* (cont.)
Study Study Design, Outcomes Results
Gaughan, 2000 24
Anticoagulation clinic for
outpatients receiving
warfarin for atrial
fibrillation (managed by
nurse practitioner using
warfarin dosing nomogram)
Radley, 1995 27
Performance of pharmacistrun
hospital-based
outpatient anticoagulation
clinic in England compared
with historical control
(management by rotating
physician trainees)
Rivey, 1993 22
Pharmacy-managed
inpatient anticoagulation
service (using weight-based
heparin protocol) for
medicine inpatients
compared with older fixeddose
protocol without any
active management by
pharmacists
Retrospective before-after
analysis (Level 3)
Percentage of patients in the
desired range for
anticoagulation (Level 2)
was evaluated as a
secondary outcome
Retrospective before-after
analysis (Level 3)
Proportions of INR
measurements “in” or “out”
of the therapeutic range
Before-after analysis
(Level 3)
Time to therapeutic PTT
(Level 2)
95
Minor increase in percentage of
patients with INR in desired range:
53.7% vs. 49.1% (p

Table 9.3. Outpatient self-management using home testing devices and dosing
nomograms*
Study Study Design, Outcomes Results
Cromheecke, 2000 38
Oral anticoagulation selfmanagement
with home
monitoring and dose
adjustment compared with
anticoagulation clinic
(Netherlands)
Sawicki, 1999 39
Oral anticoagulation selfmanagement
with home
monitoring and dose
adjustment compared with
routine care (Germany)
White, 1989 40
Oral anticoagulation selfmanagement
with home
monitoring and dose
adjustment compared with
anticoagulation clinic
(United States)
Watzke, 2000 41
Self-management
compared with
anticoagulation clinic
(Austria)
Randomized trial with
crossover comparison
(Level 1)
Adequacy of
anticoagulation (Level 2)
Single blind, multicenter
randomized controlled
trial (Level 1)
Adequacy of
anticoagulation (Level 2)
Randomized prospective
comparison (Level 1)
Adequacy of
anticoagulation (Level 2)
Prospective cohort
comparison (Level 2)
Various measures of
adequacy of
anticoagulation (Level 2)
* INR indicates international normalized ratio.
96
Percent of self-managed
measurements within 0.5 INR units of
therapeutic target did not differ (55%
vs. 49%, p=0.06). However, 29
patients (60%) during selfmanagement
spent >50% of time in
target range, compared with 25 (52%)
during clinic management (p

References
1. Bates DW, Cullen DJ, Laird N, Petersen LA, Small SD, Servi D, et al. Incidence of adverse
drug events and potential adverse drug events. Implications for prevention. ADE
Prevention Study Group. JAMA. 1995;274:29-34.
2. Cohen MR, Anderson RW, Attilio RM, Green L, Muller RJ, Pruemer JM. Preventing
medication errors in cancer chemotherapy. Am J Health Syst Pharm. 1996;53:737-746.
3. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, et al. Incidence
of adverse events and negligence in hospitalized patients. Results of the Harvard Medical
Practice Study I. N Engl J Med. 1991;324:370-376.
4. Leape LL, Kabcenell AI, Berwick DM. Reducing Adverse Drug Events. Boston: Institute
for Healthcare Improvement; 1998.
5. Massachusetts Coalition for the Prevention of Medical Errors. MHA best practice
recommendations to reduce medication errors. Available at:
www.mhalink.org/mcpme/mcpme_publications.htm. Accessed, 2001.
6. Rivey MP, Wood RD, Allington DR, Stratton TP, Erickson CC, Stenson TA. Pharmacymanaged
protocol for warfarin use in orthopedic surgery patients. Am J Health Syst Pharm.
1995;52:1310-1316.
7. Ansell J, Holden A, Knapic N. Patient self-management of oral anticoagulation guided by
capillary (fingerstick) whole blood prothrombin times. Arch Intern Med. 1989;149:2509-
2511.
8. Hasenkam JM, Knudsen L, Kimose HH, Gronnesby H, Attermann J, Andersen NT, et al.
Practicability of patient self-testing of oral anticoagulant therapy by the international
normalized ratio (INR) using a portable whole blood monitor. A pilot investigation.
Thromb Res. 1997;85:77-82.
9. Sunderji R, Campbell L, Shalansky K, Fung A, Carter C, Gin K. Outpatient selfmanagement
of warfarin therapy: a pilot study. Pharmacotherapy. 1999;19:787-793.
10. Point-of-care prothrombin time measurement for professional and patient self-testing use.
A multicenter clinical experience. Oral Anticoagulation Monitoring Study Group. Am J
Clin Pathol. 2001;115:288-296.
11. Prothrombin measurement using a patient self-testing system. Oral Anticoagulation
Monitoring Study Group. Am J Clin Pathol. 2001;115:280-287.
12. Hull RD, Raskob GE, Hirsh J, Jay RM, Leclerc JR, Geerts WH, et al. Continuous
intravenous heparin compared with intermittent subcutaneous heparin in the initial
treatment of proximal-vein thrombosis. N Engl J Med. 1986;315:1109-1114.
13. Landefeld CS, Beyth RJ. Anticoagulant-related bleeding: clinical epidemiology, prediction,
and prevention. Am J Med. 1993;95:315-328.
14. Evans RS, Pestotnik SL, Classen DC, Horn SD, Bass SB, Burke JP. Preventing adverse
drug events in hospitalized patients. Ann Pharmacother. 1994;28:523-527.
15. Becker D. Commentary on "Self-management of long-term oral anticoagulation was as
effective as specialist anticoagulation clinic management.". ACP Journal Club.
2001;134:62.
16. Raschke RA, Reilly BM, Guidry JR, Fontana JR, Srinivas S. The weight-based heparin
dosing nomogram compared with a "standard care" nomogram. A randomized controlled
trial. Ann Intern Med. 1993;119:874-881.
97

17. Elliott CG, Hiltunen SJ, Suchyta M, Hull RD, Raskob GE, Pineo GF, et al. Physicianguided
treatment compared with a heparin protocol for deep vein thrombosis. Arch Intern
Med. 1994;154:999-1004.
18. Cruickshank MK, Levine MN, Hirsh J, Roberts R, Siguenza M. A standard heparin
nomogram for the management of heparin therapy. Arch Intern Med. 1991;151:333-337.
19. Hollingsworth JA, Rowe BH, Brisebois FJ, Thompson PR, Fabris LM. The successful
application of a heparin nomogram in a community hospital. Arch Intern Med.
1995;155:2095-2100.
20. Phillips WS, Smith J, Greaves M, Preston FE, Channer KS. An evaluation and
improvement program for inpatient anticoagulant control. Thromb Haemost. 1997;77:283-
288.
21. Brown G, Dodek P. An evaluation of empiric vs. nomogram-based dosing of heparin in an
intensive care unit. Crit Care Med. 1997;25:1534-1538.
22. Rivey MP, Peterson JP. Pharmacy-managed, weight-based heparin protocol. Am J Hosp
Pharm. 1993;50:279-284.
23. Nemeth JS, Marxen TL, Piltz GW. Weight-based protocol for improving heparin therapy.
Am J Health Syst Pharm. 1996;53:1164-1166.
24. Gaughan GL, Dolan C, Wilk-Rivard E, Geary G, Libbey R, Gilman MA, et al. Improving
management of atrial fibrillation and anticoagulation in a community hospital. Jt Comm J
Qual Improv. 2000;26:18-28.
25. Chiquette E, Amato MG, Bussey HI. Comparison of an anticoagulation clinic with usual
medical care: anticoagulation control, patient outcomes, and health care costs. Arch Intern
Med. 1998;158:1641-1647.
26. Lee YP, Schommer JC. Effect of a pharmacist-managed anticoagulation clinic on warfarinrelated
hospital readmissions. Am J Health Syst Pharm. 1996;53:1580-1583.
27. Radley AS, Hall J, Farrow M, Carey PJ. Evaluation of anticoagulant control in a
pharmacist operated anticoagulant clinic. J Clin Pathol. 1995;48:545-547.
28. Pell JP, McIver B, Stuart P, Malone DN, Alcock J. Comparison of anticoagulant control
among patients attending general practice and a hospital anticoagulant clinic. Br J Gen
Pract. 1993;43:152-154.
29. Hamby L, Weeks WB, Malikowski C. Complications of warfarin therapy: causes, costs,
and the role of the anticoagulation clinic. Eff Clin Pract. 2000;3:179-184.
30. Ansell JE, Hughes R. Evolving models of warfarin management: anticoagulation clinics,
patient self-monitoring, and patient self-management. Am Heart J. 1996;132:1095-1100.
31. Bussey HI, Rospond RM, Quandt CM, Clark GM. The safety and effectiveness of longterm
warfarin therapy in an anticoagulation clinic. Pharmacotherapy. 1989;9:214-219.
32. Wilt VM, Gums JG, Ahmed OI, Moore LM. Outcome analysis of a pharmacist-managed
anticoagulation service. Pharmacotherapy. 1995;15:732-739.
33. Cortelazzo S, Finazzi G, Viero P, Galli M, Remuzzi A, Parenzan L, et al. Thrombotic and
hemorrhagic complications in patients with mechanical heart valve prosthesis attending an
anticoagulation clinic. Thromb Haemost. 1993;69:316-320.
34. Garabedian-Ruffalo SM, Gray DR, Sax MJ, Ruffalo RL. Retrospective evaluation of a
pharmacist-managed warfarin anticoagulation clinic. Am J Hosp Pharm. 1985;42:304-308.
35. Hamilton GM, Childers RW, Silverstein MD. Does clinic management of anticoagulation
improve the outcome of prosthetic valve patients? Clin res. 1985;33:832A.
36. Cohen IA, Hutchison TA, Kirking DM, Shue ME. Evaluation of a pharmacist-managed
anticoagulation clinic. J Clin Hosp Pharm. 1985;10:167-175.
98

37. Ellis RF, Stephens MA, Sharp GB. Evaluation of a pharmacy-managed warfarinmonitoring
service to coordinate inpatient and outpatient therapy. Am J Hosp Pharm.
1992;49:387-394.
38. Cromheecke ME, Levi M, Colly LP, de Mol BJ, Prins MH, Hutten BA, et al. Oral
anticoagulation self-management and management by a specialist anticoagulation clinic: a
randomised cross-over comparison. Lancet. 2000;356:97-102.
39. Sawicki PT. A structured teaching and self-management program for patients receiving
oral anticoagulation: a randomized controlled trial. Working Group for the Study of Patient
Self-Management of Oral Anticoagulation. JAMA. 1999; 281:145-150.
40. White RH, Becker DM, Gunther-Maher MG. Outpatient use of a portable international
normalized ratio/prothrombin time monitor. South Med J. 1994;87:206-210.
41. Watzke HH, Forberg E, Svolba G, Jimenez-Boj E, Krinninger B. A prospective controlled
trial comparing weekly self-testing and self-dosing with the standard management of
patients on stable oral anticoagulation. Thromb Haemost. 2000;83:661-665.
42. Christensen TD, Attermann J, Pilegaard HK, Andersen NT, Maegaard M, Hasenkam JM.
Self-management of oral anticoagulant therapy for mechanical heart valve patients. Scand
Cardiovasc J. 2001;35:107-113.
43. Heidinger KS, Bernardo A, Taborski U, Muller-Berghaus G. Clinical outcome of selfmanagement
of oral anticoagulation in patients with atrial fibrillation or deep vein
thrombosis. Thromb Res. 2000;98:287-293.
44. Hasenkam JM, Kimose HH, Knudsen L, Gronnesby H, Halborg J, Christensen TD, et al.
Self management of oral anticoagulant therapy after heart valve replacement. Eur J
Cardiothorac Surg. 1997;11:935-942.
45. Ansell JE, Patel N, Ostrovsky D, Nozzolillo E, Peterson AM, Fish L. Long-term patient
self-management of oral anticoagulation. Arch Intern Med. 1995;155:2185-2189.
46. White RH, McCurdy SA, von Marensdorff H, Woodruff DE, Jr., Leftgoff L. Home
prothrombin time monitoring after the initiation of warfarin therapy. A randomized,
prospective study. Ann Intern Med. 1989;111:730-737.
47. Ansell JE, Hamke AK, Holden A, Knapic N. Cost effectiveness of monitoring warfarin
therapy using standard versus capillary prothrombin times. Am J Clin Pathol. 1989;91:587-
589.
48. Taborski U, Wittstamm FJ, Bernardo A. Cost-effectiveness of self-managed anticoagulant
therapy in Germany. Semin Thromb Hemost. 1999;25:103-107.
49. Lafata JE, Martin SA, Kaatz S, Ward RE. Anticoagulation clinics and patient self-testing
for patients on chronic warfarin therapy: A cost-effectiveness analysis. J Thromb
Thrombolysis. 2000;9(Suppl 1):S13-S19.
99

100

Chapter 10. Unit-Dose Drug Distribution Systems
Michael D. Murray, PharmD, MPH
Purdue University School of Pharmacy
Kaveh G. Shojania, MD
University of California, San Francisco School of Medicine
Background
Medication errors are especially likely when health professionals engage in multiple
tasks within a short time span. This situation occurs repeatedly in hospitals when pharmacists
and technicians load unit-dose carts, and when nurses administer medications. Unit-dose carts
are prepared daily, often manually, by technicians and then checked by pharmacists. These carts,
containing thousands of patient-specific dosages of drugs, are sent to the wards daily, for nurses
to administer medications to patients. Dosing frequencies vary widely, ranging from regular
intervals around the clock to “stat” doses given to control acute pain or other symptoms.
Medication administration alone is an enormous task for nurses, and one in which they are
repeatedly interrupted. It is not surprising that the administration phase of medication use is
particularly vulnerable to error. 1
Unit-dose dispensing of medication was developed in the 1960s to support nurses in
medication administration and reduce the waste of increasingly expensive medications. Most of
the investigations of medication errors and unit-dose dispensing took place from 1970 to 1976.
Now, unit-dose dispensing of medications is a standard of practice at hospitals in the United
States. This chapter reviews the evidence supporting this practice.
Practice Description
In unit-dose dispensing, medication is dispensed in a package that is ready to administer
to the patient. 2 It can be used for medications administered by any route, but oral, parenteral, and
respiratory routes are especially common. When unit-dose dispensing first began, hospital
pharmacies equipped themselves with machines that packaged and labeled tablets and capsules,
one pill per package. They also purchased equipment for packaging liquids in unit-doses. As the
popularity of this packaging increased, the pharmaceutical industry began prepackaging pills in
unit-of-use form. Many hospitals now purchase prepackaged unit-dose medications. However, it
is still common for hospital pharmacies to purchase bulk supplies of tablets and capsules from
manufacturers and repackage them in the central pharmacy into unit-dose packages. 2 It is
important to note that hospitals vary in the proportion of their wards covered by a unit-dose
system.
There are many variations of unit-dose dispensing. As just one example, when physicians
write orders for inpatients, these orders are sent to the central pharmacy (by pharmacists, nurses,
other personnel, or computer). Pharmacists verify these orders and technicians place drugs in
unit-dose carts. The carts have drawers in which each patient’s medications are placed by
pharmacy technicians—one drawer for each patient. The drawers are labeled with the patient’s
name, ward, room, and bed number. Before the carts are transported to the wards, pharmacists
check each drawer’s medications for accuracy. Sections of each cart containing all medication
drawers for an entire nursing unit often slide out and can be inserted into wheeled medication
carts used by nurses during their medication administration cycles. A medication administration
recording form sits on top of the cart and is used by the nurse to check-off and initial the time of
each administration of each medication. The next day, the carts are retrieved from the wards and
101

eplaced by a fresh and updated medication supply. Medications that have been returned to the
central pharmacy are credited to the patient’s account.
A 1999 national survey of drug dispensing and administration practices indicated that
three-fourths of responding hospitals had centralized pharmacies, 77% of which were not
automated. 2 Larger hospitals and those affiliated with medical schools were more likely to have
some component of decentralized pharmacy services. About half of the surveyed hospitals
reported drug distribution “systems” that bypassed the pharmacy, including hospitals that
reported using floor stocks, borrowing other patients’ medications, and hidden drug supplies.
Studies often compare unit-dose dispensing to a ward stock system. In this system, nurses
order drugs in bulk supplies from the pharmacy; the drugs are stored in a medication room on the
ward. Nurses prepare medication cups for each patient during medication administration cycles.
The correct number of pills must be taken out of the correct medication container for each cycle
and taken to the patient for administration. Liquids must be poured by the nurse from the
appropriate bottle and each dose carefully measured. Nurses are responsible for any necessary
labeling. Any medications taken from stock bottles and not administered to patients are generally
disposed of.
Prevalence and Severity of the Target Safety Problem
The targets of the safety problem for unit-dosing are drug dispensing 3 and
administration. 4,5 Improving these stages probably carries the greatest opportunity to reduce
medication errors.
Bates et al 6 identified 530 medical errors in 10,070 written orders for drugs (5.3
errors/100 orders) on 3 medical units observed for 51 days. Of the 530 errors, 5 (0.9%) resulted
in an adverse drug event. The most common reason for an error was a missing dose of
medication, which occurred in 53% of orders. In a systems analysis of 334 errors causing 264
adverse drug events over 6 months in 2 tertiary care hospitals, 130 errors (39%) resulted from
physician ordering, 40 (12%) involved transcription and verification, 38 (11%) reflected
problems with pharmacy dispensing, and 126 (38%) were from nursing administration. 4 In other
words, 164 (49%) of the errors in the above-cited study 4 were in the dispensing and
administration stages. In further research, the investigators found that errors resulting in
preventable adverse drug events were more than likely to be those in the administration stage
(34%) than those in the dispensing stage (4%) of the medication use process. 1
Opportunities for Impact
Because unit-dose dispensing now constitutes a standard for approval by the Joint
Commission for Accreditation of Healthcare Organizations (JCAHO) 7,8 and is closely linked to
the increasingly common use of automated dispensing devices (see Chapter 11), there is likely
little opportunity for further implementation of this practice in US hospitals. In a 1994 survey of
pharmacy directors, 92% of acute care hospitals reported using unit-dose dispensing. 7 Use of
unit-dose dispensing is extremely common on general medical and surgical wards, but less so in
other locations such as intensive care units, operating rooms, and emergency departments. In
these areas, bulk medication stock systems are still found. In a 1999 survey of pharmacy
directors, 80% reported that unit-dose dispensing was used for 75% or more of oral doses, and
52% of injectable medications dispensed in their hospitals. 9
102

Study Designs
The 5 studies meeting inclusion criteria for this review (Table 10.1) included 4 Level 3
studies, and one prospectively designed before-after observation (Level 2 design). We excluded
the following studies for the reasons outlined:
• Read 10 conducted a study of a unit-dose system applied to respiratory therapy
solutions that was focused primarily on measures of efficiency.
•
Volume/concentration errors in preparing respiratory solutions were also
discussed, but the method of detection was not stated nor were specific error
data provided for the study period.
Reitberg et al 11 conducted a prospective before-after study with a concurrent
comparison floor as a control (Level 2 design) in which medication errors
represented the main outcomes (Level 2). During period 1 when both wards
used the ward stock system, total errors were 4.7% on the ward that remained
on ward stock and 34.2% on the ward that used the modified dispensing
system. During period 2, the ward stock system had 5.1% error, while the
intervention ward (unit-dose) exhibited a significantly increased 18% error
rate. Excluding administration-time errors resulted in error rates of 5.1%
(conventional system) and 4.8% (unit-dose), which are not significantly
different. The greater number of errors on the modified dispensing ward
before and after the intervention were attributed by the authors to factors other
than the dispensing systems. Because of this problem, and the fact that this
study took place in a skilled nursing facility rather than an acute care hospital,
we excluded this study.
• Shultz et al 12 conducted a prospective before-after study (Level 2) involving a
somewhat complicated comparison. In the baseline period, approximately
50% of wards used a unit-dose system. During the study period, some wards
switched to a unit-dose system in which nurses still administered the
medications, while other wards adopted an experimental system, in which
pharmacists and pharmacy technicians handled all aspects of the mediation
dispensing and administration process. The authors report an error rate of
0.64% for complete unit-dose plus pharmacy technician medication system
compared to 5.3% in the more conventional unit-dose system in which nurses
continue to administer medications (p

Lastly, we were unable to obtain one of the original studies of the unit-dose system
within the timeline of the project. The data from this study appears to have been published only
as a technical report. 15 Other publications related to this study and which we were able to
obtain 16-18 did not provide sufficient detail on the aspects of the study relating to medication
errors to permit abstraction or inclusion in this chapter. The published reports of a study
conducted in a private hospital similarly did not contain sufficient detail about the study methods
or results to permit abstraction and inclusion. 19,20
Study Outcomes
Studies reported errors measured by direct observation using a methodology that was first
described by Barker. 21 All of these studies involved Level 2 outcomes.
Evidence for Effectiveness of the Practice
Though the practice of unit-dose dispensing is generally well accepted and has been
widely implemented, the evidence for its effectiveness is modest. Most of the published studies
reported reductions in medication errors of omission and commission with unit-dose dispensing
compared with alternative dispensing systems such as ward stock systems. One exception was
the international study by Dean, 22 which compared a United States hospital using unit-dose and
transcription of medication orders with a United Kingdom hospital using ward stock and no
transcription of orders. The results of this study indicated that the United Kingdom hospital had
less than half the errors of the United States hospital. The study groups were often difficult to
compare because of differing cultures, hospitals, or nursing care units. Moreover, cointerventions
undoubtedly confounded the results. Each study is summarized in Table 10.1.
Potential for Harm
Unit-dosing shifts the effort and distraction of medication processing, with its potential
for harm, from the nursing ward to central pharmacy. It increases the amount of time nurses have
to do other tasks but increases the volume of work within the pharmacy. Like the nursing units,
central pharmacies have their own distractions that are often heightened by the unit-dose
dispensing process itself, and errors do occur. 3
Overall, unit-dose appears to have little potential for harm. The results of most of the
Level 2 observational studies seem to indicate that it is safer than other forms of institutional
dispensing. However, the definitive study to determine the extent of harm has not yet been
conducted.
A major advantage of unit-dose dispensing is that it brings pharmacists into the
medication use process at another point to reduce error. 3 Yet, as pointed out in the Practice
Description above, about half of the hospitals in a national survey bypass pharmacy involvement
by using floor stock, borrowing patients’ medications, and hiding medication supplies. 2
Costs and Implementation
The cost considerations of unit-dose dispensing are mainly a trade-off between pharmacy
and nursing personnel. The pharmacy personnel involved are mainly technicians who load unitdose
ward carts for the pharmacists to check and may package some medications that are
commercially unavailable in unit-dose form. The pharmacist must verify that the correct
medication and dosage of each medication is sent to the ward for the nurse to administer.
Nursing time to maintain a drug inventory is reduced, allowing more time for other nursing
activities. A variable cost of unit-dose dispensing is the cost of equipment and supplies to those
104

hospitals that wish to do much of the packaging themselves instead of purchasing medications
pre-packaged as unit-doses.
Comment
The studies evaluating the practice of unit-dosing contain important methodologic
problems and, although yielding somewhat heterogeneous results, are overall relatively
consistent in showing a positive impact on error reduction. In contrast to other practices related
to medication use, none of the studies evaluated Level 1 outcomes, such as actual adverse drug
events. Nonetheless, unit-dose dispensing or some form of automated dispensing of unit-doses
(see Chapter 11) has become ubiquitous in American hospitals and a standard of care in the
delivery of pharmacy services. Consequently, it is unlikely that more rigorous studies could now
be conducted.
105

Table 10.1. Studies evaluating the impact of unit-dose dispensing on medication errors*
Study Study Design, Outcomes† Results: Error Rates (95% CI)
Hynniman,
1970 23
Means,
1975 13
Simborg,
1975 14 ‡
Schnell,
1976 24
Dean,
1995 22
Taxis,
1998 25
Cross-sectional comparison between
study hospital and non-randomly
selected “comparison” hospitals (Level
3)
Errors of commission and omission
(Level 2) among doses ordered
Cross-sectional comparison of 2 wards
within a single hospital over a 60-day
period (Level 3)
Errors of commission (Level 2) among
doses administered during randomly
chosen observation periods
Prospective before-after study (Level
2) at four Canadian hospitals
Errors observed during medication
preparation and administration (Level
2)
Cross-sectional comparison (Level 3) of
US and UK hospitals with different
pharmacy distribution systems
Errors observed during medication
administration (Level 2)
Cross-sectional comparison (Level 3) of
2 hospitals in Germany and one hospital
in the UK
Errors observed during medication
administration
106
Unit-dose system: 3.5% (3.1-4.0%)
Conventional distribution systems at 4 hospitals:
8.3% (7.1-9.7%)
9.9% (8.0-12.2%)
11.4% (9.9-13.2%)
20.6% (18.4-22.9%)
Unit-dose ward: 1.6% (1.0-2.5%)
Multi-dose ward: 7.4% (6.1-8.9%)§
Before vs. after implementation of unit-dose
system: 37.2 vs. 38.5%;
42.9 vs. 23.3%;
20.1% vs. 7.8%;
38.5% vs. 23.1%
84 errors among 2756 observations in UK
hospital using traditional ward stock system:
3.0% (2.4-3.7%)
63 errors among 919 observations in US hospital
using unit-doses and automated dispensing: 6.9%
(5.2-8.5%)
Absolute difference: 3.9% (2.1-5.7%)
UK hospital using traditional ward stock system:
8.0% (6.2-9.8%)
German hospital using traditional ward stock
system: 5.1% (4.4-5.8)
German hospital using unit-dose system: 2.4%
(2.0-2.8%)
Omission was the most common type of error
* CI indicates confidence interval.
† Errors of commission include administration of wrong dose or wrong or unordered drug, whereas errors
of omission include missed doses for inclusion in a patient’s unit-dose drawer or a dose not
administered.
‡ As outlined in the text, the similarities in study setting, time, design and results suggest that these 2
references contain data from the same study; information from these references was therefore combined
and treated as a single study.
§ The 95% CIs shown in the table were calculated using the reported data: 20 errors in 1234 observed
doses on the unit-dose ward vs. 105 errors in 1428 observed doses on the multidose ward.

When wrong time errors were omitted, the above results changed so that the change to a unit-dose was
associated with a significant increase in errors at the first hospital, a non-significant decrease in errors at
the second hospital, and significant decreases in errors at the other two hospitals.
107

References
1. Bates DW, Cullen DJ, Laird N, Petersen LA, Small SD, Servi D, et al. Incidence of adverse
drug events and potential adverse drug events: Implications for prevention. JAMA.
1995;274:29-34.
2. Ringold DJ, Santell JP, Schneider PJ, Arenberg S. ASHP national survey of pharmacy
practice in acute care settings: prescribing and transcribing–1998. American Society of
Health-System Pharmacists. Am J Health Syst Pharm. 1999;56:142-157.
3. Facchinetti NJ, Campbell GM, Jones DP. Evaluating dispensing error detection rates in a
hospital pharmacy. Med Care. 1999;37:39-43.
4. Leape L, Bates DW, Cullen DJ, Cooper J, Demonaco HJ, Gallivan T, et al. Systems
analysis of adverse drug events. JAMA. 1995;274:35-43.
5. Tissot E, Cornette C, Demoly P, Jacquet M, Barale F, Capellier G. Medication errors at the
administration stage in an intensive care unit. Intensive Care Med. 1999;25:353-359.
6. Bates DW, Boyle DL, Vander Vliet MB, Schneider J, Leape L. Relationship between
medication errors and adverse drug events. J Gen Intern Med. 1995;10:199-205.
7. Santell JP. ASHP national survey of hospital-based pharmaceutical services–1994. Am J
Health Syst Pharm. 1995;52:1179-1198.
8. Godwin HN, Scott BE. Institutional patient care. In: Gennaro AR, editor. Remington: The
Science and Practice of Pharmacy. Baltimore: Lippincott Williams & Wilkins; 2000:1911-
31.
9. Ringold DJ, Santell JP, Schneider PJ. ASHP national survey of pharmacy practice in acute
care settings: dispensing and administration–1999. Am J Health Syst Pharm.
2000;57:1759-1775.
10. Read OK. Unit dose packaging of respiratory therapy solutions. Am J Health Syst Pharm.
1975;32:49-51.
11. Reitberg DP, Miller RJ, Bennes JF. Evaluation of two concurrent drug-delivery systems in
a skilled-nursing facility. Am J Health Syst Pharm. 1982;39:1316-1320.
12. Shultz SM, White SJ, Latiolais CJ. Medication errors reduced by unit-dose. Hospitals.
1973;47:107-112.
13. Means BJ, Derewicz HJ, Lamy PP. Medication errors in a multidose and computer-based
unit dose drug distribution system. Am J Hosp Pharm. 1975;32:186-191.
14. Simborg DW, Derewicz HJ. A highly automated hospital medication system: Five year's
experience and evaluation. Ann Intern Med. 1975;83:342-346.
15. A study of patient care involving a unit dose system. Final report: University of Iowa;
1966. U.S.P.H.S. Grant # HM-00328-01.
16. Greth PA, Tester WW, Black HJ. Decentralized pharmacy operations utilizing the unit
dose concept. II. Drug information services and utilization in a decentralized pharmacy
substation. Am J Hosp Pharm. 1965;22:558-563.
17. Black HJ, Tester WW. Decentralized pharmacy operations utilizing the unit dose concept.
3. Am J Hosp Pharm. 1967;24:120-129.
18. Black HJ, Tester WW. The unit dose drug-distribution system. J Iowa Med Soc.
1968;58:1174-1176.
19. Slater WE, Hripko JR. The unit-dose system in a private hospital. 1. Implementation. Am J
Hosp Pharm. 1968;25:408-417.
20. Slater WE, Hripko JR. The unit-dose system in a private hospital. II. Evaluation. Am J
Hosp Pharm. 1968;25:641-648.
108

21. Barker KN, McConnell WE. The problems of detecting medication errors in hospitals. Am
J Hosp Pharm. 1962;19:361-369.
22. Dean BS, Allan EL, Barber ND, Barker KN. Comparison of medication errors in an
American and British hospital. Am J Health Syst Pharm. 1995;52:2543-2549.
23. Hynniman CE, Conrad WF, Urch WA, Rudnick BR, Parker PF. A comparison of
medication errors under the University of Kentucky unit dose system and traditional drug
distribution systems in four hospitals. Am J Hosp Pharm. 1970;27:803-814.
24. Schnell BR. A study of unit-dose drug distribution in four Canandian hospitals. Can J
Hosp Pharm. 1976;29:85-90.
25. Taxis K, Dean B, Barber N. Hospital drug distribution systems in the UK and Germany - a
study of medication errors. Pharm World Sci. 1998;21:25-31.
109

110

Chapter 11. Automated Medication Dispensing Devices
Michael D. Murray, PharmD, MPH
Purdue University School of Pharmacy
Background
In the 1980s, automated dispensing devices appeared on the scene, a generation after the
advent of unit-dose dispensing (Chapter 11). The invention and production of these devices
brought hopes of reduced rates of medication errors, increased efficiency for pharmacy and
nursing staff, ready availability of medications where they are most often used (the nursing unit
or inpatient ward), and improved pharmacy inventory and billing functions. 1-4 Although the
capacity of such systems to contribute to patient safety appears great, surprisingly few studies
have evaluated the clinical impact of these devices.
Practice Description
Automated dispensing systems are drug storage devices or cabinets that electronically
dispense medications in a controlled fashion and track medication use. Their principal advantage
lies in permitting nurses to obtain medications for inpatients at the point of use. Most systems
require user identifiers and passwords, and internal electronic devices track nurses accessing the
system, track the patients for whom medications are administered, and provide usage data to the
hospital’s financial office for the patients’ bills.
These automated dispensing systems can be stocked by centralized or decentralized
pharmacies. Centralized pharmacies prepare and distribute medications from a central location
within the hospital. Decentralized pharmacies reside on nursing units or wards, with a single
decentralized pharmacy often serving several units or wards. These decentralized pharmacies
usually receive their medication stock and supplies from the hospital’s central pharmacy.
More advanced systems provide additional information support aimed at enhancing
patient safety through integration into other external systems, databases, and the Internet. Some
models use machine-readable code for medication dispensing and administration. Three types of
automated dispensing devices were analyzed in the studies reviewed here, the McLaughlin
dispensing system, the Baxter ATC-212 dispensing system, and the Pyxis Medstation Rx. Their
attributes are described below.
• The McLaughlin dispensing system 5 includes a bedside dispenser, a
programmable magnetic card, and a pharmacy computer. It is a locked system
that is loaded with the medications prescribed for a patient. At the appropriate
dosing time, the bedside dispenser drawer unlocks automatically to allow a
dose to be removed and administered. A light above the patient’s door
illuminates at the appropriate dosing time. Only certain medications fit in the
compartmentalized cabinet (such as tablets, capsules, small pre-filled
syringes, and ophthalmic drops).
• The Baxter ATC-212 dispensing system 6 uses a microcomputer to pack unitdose
tablets and capsules for oral administration. It is usually installed at the
pharmacy. Medications are stored in calibrated canisters that are designed
specifically for each medication. Canisters are assigned a numbered location,
which is thought to reduce mix-up errors upon dispensing. When an order is
sent to the microcomputer, a tablet is dispensed from a particular canister. The
111

drug is ejected into a strip-packing device where it is labeled and hermetically
sealed.
• The Pyxis Medstation, Medstation Rx, and Medstation Rx 1000 are automated
dispensing devices kept on the nursing unit. 7-9 These machines are often
compared to automatic teller machines (ATMs). The Medstation interfaces
with the pharmacy computer. Physicians’ orders are entered into the pharmacy
computer and then transferred to the Medstation where patient profiles are
displayed to the nurse who accesses the medications for verified orders. Each
nurse is provided with a password that must be used to access the Medstation.
Pharmacists or technicians keep these units loaded with medication. Charges
are made automatically for drugs dispensed by the unit. Earlier models had
sufficient memory to contain data for about one week, and newer models can
store data for longer periods.
Studies reviewed did not include the automated dispensing systems manufactured by
Omnicell, which produces point-of-use systems that can be integrated into a hospital’s
information system. 10 Omnicell systems are also capable of being integrated into external
support systems that support machine-readable code, drug information services, and medication
error reporting systems.
Prevalence and Severity of the Target Safety Problem
Medication errors within hospitals occur with 2% to 17% of doses ordered for
inpatients. 5,7,11-14 It has been suggested that the rate of inpatient medication errors is one per
patient per inpatient day. 15 The specific medication errors targeted by automated dispensing
systems are those related to drug dispensing and administration. Even with the use of unit-doses
(see Chapter 11) errors still occur at the dispensing 16 and administration stages 3,17 of the
medication use process. For instance, in one large study of 530 medical errors in 10,070 written
orders for drugs (5.3 errors/100 orders), 18 pharmacy dispensing accounted for 11% of errors and
nursing administration 38%. 3
Opportunities for Impact
Automated dispensing devices have become increasingly common either to supplement
or replace unit-dose distribution systems in an attempt to improve medication availability,
increase the efficiency of drug dispensing and billing, and reduce errors. A 1999 national survey
of drug dispensing and administration practices indicated that 38% of responding hospitals used
automated medication dispensing units and 8.2% used machine-readable coding with
dispensing. 19 Three-fourths of respondents stated that their pharmacy was centralized and of
these centralized pharmacies, 77% were not automated. Hospitals with automated centralized
pharmacies reported that greater than 50% of their inpatient doses were dispensed via centralized
automated systems. Half of all responding hospitals used a decentralized medication storage
system. One-third of hospitals with automated storage and dispensing systems were linked to the
pharmacy computer. Importantly, about half of the surveyed hospitals reported drug distributions
that bypassed the pharmacy including floor stock, borrowing patients’ medications, and hidden
drug supplies.
Study Designs
There were no true randomized trials. One crossover study of the McLaughlin dispensing
system randomized nurses to work with the intervention medication system or the control
112

system. 5 We classified this as a Level 2 study, since, from the patient perspective, the design is
that of a non-randomized trial. Other studies included in this review consisted of retrospective
observational studies with before-after 6-8 or cross-sectional design 11 (Level 3). The reviewed
studies described dispensing systems for orally administered medications, and were published
between 1984 and 1995 (see Table 12.1).
Study Outcomes
All studies measured rates of medication errors (Level 2 outcome). Four studies 5,7,8,11
detected errors by direct observation using a methodology that was first described by Barker. 5
Direct observation methods have been criticized because of purported Hawthorne effect (bias
involving changed behavior resulting from measurements requiring direct observation of study
subjects). However, proponents of the method state that such effects are short-lived, dissipating
within hours of observation. 15 Dean and Barber have recently demonstrated the validity and
reliability of direct observational methods to detect medication administration errors. 20 Another
study, a Level 3 design, determined errors by inspecting dispensed drugs. 6
Evidence for Effectiveness of the Practice
The evidence provided by the limited number of available, generally poor quality studies
does not suggest that automated dispensing devices reduce medication errors. There is also no
evidence to suggest that outcomes are improved with the use of these devices. Most of the
published studies comparing automated devices with unit-dose dispensing systems report
reductions in medication errors of omission and scheduling errors with the former. 7,9 The studies
suffer from multiple problems with confounding, as they often compare hospitals or nursing care
units that may differ in important respects other than the medication distribution system.
Potential for Harm
Human intervention may prevent these systems from functioning as designed.
Pharmacists and nurses can override some of the patient safety features. When the turn around
time for order entry into the automated system is prolonged, nurses may override the system
thereby defeating its purpose. Furthermore, the automated dispensing systems must be refilled
intermittently to replenish exhausted supplies. Errors can occur during the course of refilling
these units or medications may shift from one drawer or compartment to another causing
medication mix-ups. Either of these situations can slip past the nurse at medication
administration.
The results of the study of the McLaughlin dispensing system indicated that though
overall errors were reduced compared to unit-dose (10.6% vs. 15.9%), errors decreased for 13 of
20 nurses but increased for the other 7 nurses. 5 In a study of Medstation Rx vs. unit-dose, 8 errors
decreased in the cardiovascular surgery unit, where errors were recorded by work measurement
observations. However, errors increased over 30% in 6 of 7 nurses after automated dispensing
was installed in the cardiovascular intensive care unit, where incident reports and medication
error reports were both used for ascertaining errors, raising the question of measurement bias.
Finally, in a study primarily aimed at determining differences in errors for ward and unit-dose
dispensing systems, 11 a greater error prevalence was found for medications dispensed using
Medstation Rx compared with those dispensed using unit-dose or non-automated floor stock
(17.1% vs. 5.4%).
113

Costs and Implementation
The cost of automated dispensing mainly involves the capital investment of renting or
purchasing equipment for dispensing, labeling, and tracking (which often is done by computer).
A 1995 study revealed that the cost of Medstation Rx to cover 10 acute care units (330 total
beds) and 4 critical care units (48 total beds) in a large referral hospital would be $1.28 million
over 5 years. Taking into account costs saved from reduced personnel and decreased drug waste,
the units had the potential to save $1 million over 5 years. Most studies that examine economic
impact found a trade-off between reductions in medication dispensing time for pharmacy and
medication administration time for nursing personnel. A common complaint by nurses is long
waiting lines at Pyxis Medstations if there are not enough machines. Nurses must access these
machines using a nurse-specific password. This limited access to drugs on nursing units
decreases drug waste and pilferage.
Comment
Although the implementation of automated dispensing reduces personnel time for
medication administration and improves billing efficiency, reduction in medication errors have
not been uniformly realized. Indeed, some studies suggest that errors may increase with some
forms of automation. The results of the study of the McLaughlin Dispensing System by Barker et
al 5 showed considerable nurse-to-nurse variability in the error rate between the automated
system and conventional unit dose. Qualitative data aimed at determining the reason for this
variability would be useful. The study by Klein et al 6 indicated little difference in the accuracy of
medication cart filling by the Baxter ATC-212 (0.65%) versus filling by technicians (0.84%).
Borel and Rascati found that medication errors, largely those related to the time of
administration, were fewer after implementation of the Pyxis Medstation Rx (10.4%) compared
with the historical period (16.9%). 7 These results are consistent with a more recent study by
Shirley, that found a 31% increase in the on-time administration of scheduled doses after
installation of the Medstation Rx 1000. 9 In contrast, errors were greater after Medstation Rx in
the study by Schwarz and Brodowy, 8 increasing on 6 of 7 nursing units by more than 30%.
Finally, Dean et al found half the errors in a ward-based system without automation in the
United Kingdom (3.0%, 95% CI: 2.4-3.7%) compared with an automated unit-dose medication
distribution system in the United States (6.9%, 95% CI: 5.2-8.5%). 11
The practical limitations of the systems were illustrated by a variety of process deviations
observed by Borel and Rascati. 7 These included nurses waiting at busy administration times,
removal of doses ahead of time to circumvent waiting, and overriding the device when a dose
was needed quickly. These procedural failures emphasize an often-raised point with the
introduction of new technologies, namely that the latest innovations are not a solution for
inadequate or faulty processes or procedures. 2
Although automated dispensing systems are increasingly common, it appears they may
not be completely beneficial in their current form. Further study is needed to demonstrate the
effectiveness of newer systems such as the Omnicell automated dispensing devices. If the
standard, namely unit-dose dispensing, is to be improved, such improvements will likely derive
from robotics and informatics. To document impact of automated dispensing devices on patient
safety, studies are needed comparing unit-dose dispensing with automated dispensing devices.
Until the benefits of automated dispensing devices become clearer, the opportunities for impact
of these devices is uncertain.
114

Table 11.1. Six studies reviewing automated drug dispensing systems*
Study Study Design Study Outcomes N Results
Barker,
1984 5
Klein,
1994 6
Borel,
1995 7
Schwarz,
1995 8
Dean,
1995 11
Prospective
controlled
clinical trial
(Level 2)
Prospective
comparison of
two cohorts
(Level 2)
Prospective
before-after study
(Level 2)
Prospective
before-after study
(Level 2)
Cross-sectional
comparison
(Level 3) of US
and UK hospitals
with different
pharmacy
distribution
systems
* CI indicates confidence interval.
† Study used various denominator data.
Errors of omission and
commission among
number of ordered and
unauthorized doses.
(Level 2)
Dispensing errors in
unit-dose drawers to be
delivered to nursing
units (Level 2)
Errors observed during
medication
administrationin
medications
administered (Level 2)
Errors in medications
administered (Level 2)
Errors in medications
administered (Level 2)
115
1775 96 errors among 902 observations
(10.6%) using the McLaughlin
dispensing system vs. 139 errors
among 873 observations (15.9%)
using unit-dose dispensing (control)
7842 34 errors found among 4029 doses
(0.84%) filled manually by
technicians vs. 25 errors among 3813
doses (0.66%) filled by automated
dispensing device
1802 148 errors among 873 observations
(16.9%) before vs. 97 errors among
929 observations (10.4%) after
Medstation Rx (p

References
1. Perini VJ, Vermeulen LC, Jr. Comparison of automated medication-management systems.
Am J Hosp Pharm. 1994;51:1883-1891.
2. Murray MD. Information technology: the infrastructure for improvements to the
medication-use process. Am J Health Syst Pharm. 2000;57:565-571.
3. Guerrero RM, Nickman NA, Jorgenson JA. Work activities before and after
implementation of an automated dispensing system. Am J Health Syst Pharm.
1996;53:548-554.
4. Felkey BG, Barker KN. Technology and automation in pharmaceutical care. J Am Pharm
Assoc (Wash). 1996;NS36:309-314.
5. Barker KN, Pearson RE, Hepler CD, Smith WE, Pappas CA. Effect of an automated
bedside dispensing machine on medication errors. Am J Hosp Pharm. 1984;41:1352-1358.
6. Klein EG, Santora JA, Pascale PM, Kitrenos JG. Medication cart-filling time, accuracy,
and cost with an automated dispensing system. Am J Hosp Pharm. 1994;51:1193-1196.
7. Borel JM, Rascati KL. Effect of an automated, nursing unit-based drug-dispensing device
on medication errors . Am J Health Syst Pharm. 1995;52:1875-1879.
8. Schwarz HO, Brodowy BA. Implementation and evaluation of an automated dispensing
system. Am J Health Syst Pharm. 1995;52:823-828.
9. Shirley KL. Effect of an automated dispensing system on medication administration time.
Am J Health Syst Pharm. 1999;56:1542-1545.
10. Omnicell Automated Systems. Omnicell Homepage. Available at:
http://www.omnicell.com. Accessed May 28, 2001.
11. Dean BS, Allan EL, Barber ND, Barker KN. Comparison of medication errors in an
American and British hospital. Am J Health Syst Pharm. 1995;52:2543-2549.
12. Bates DW, Cullen DJ, Laird N, Petersen LA, Small SD, Servi D, et al. Incidence of adverse
drug events and potential adverse drug events: Implications for prevention. JAMA.
1995;274:29-34.
13. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in
hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA.
1997;277:301-306.
14. Cullen DJ, Sweitzer BJ, Bates DW, Burdick E, Edmondson A, Leape LL. Preventable
adverse drug events in hospitalized patients: a comparative study of intensive care and
general care units. Crit Care Med. 1997;25:1289-1297.
15. Allan EL, Barker KN. Fundamentals of medication error research. Am J Hosp Pharm.
1990;47:555-571.
16. Facchinetti NJ, Campbell GM, Jones DP. Evaluating dispensing error detection rates in a
hospital pharmacy. Med Care. 1999;37:39-43.
17. Tissot E, Cornette C, Demoly P, Jacquet M, Barale F, Capellier G. Medication errors at the
administration stage in an intensive care unit. Intensive Care Med. 1999;25:353-3359.
18. Bates DW, Boyle DL, Vander Vliet MB, Schneider J, Leape L. Relationship between
medication errors and adverse drug events. J Gen Intern Med. 1995;10:199-205.
19. Ringold DJ, Santell JP, Schneider PJ. ASHP national survey of pharmacy practice in acute
care settings: dispensing and administration–1999. Am J Health Syst Pharm.
2000;57:1759-1775.
20. Dean B, Barber N. Validity and reliability of observational methods for studying
medication administration errors. Am J Health Syst Pharm. 2001;58:54-59.
116

117

Section B. Infection Control
Chapter 12. Practices to Improve Handwashing Compliance
Chapter 13. Impact of Barrier Precautions in Reducing the Transmission of Serious
Nosocomial Infections
Chapter 14. Impact of Changes in Antibiotic Use Practices on Nosocomial Infections and
Antimicrobial Resistance – Clostridium Difficile and Vancomycin-resistant
Enterococcus (VRE)
Chapter 15. Prevention of Nosocomial Urinary Tract Infections
Subchapter 15.1. Use of Silver Alloy Urinary Catheters
Subchapter 15.2. Use of Suprapubic Catheters
Chapter 16. Prevention of Intravascular Catheter-Associated Infections
Subchapter 16.1. Use of Maximum Barrier Precautions during Central Venous
Catheter Insertion
Subchapter 16.2. Use of Central Venous Catheters Coated with Antibacterial or
Antiseptic Agents
Subchapter 16.3. Use of Chlorhexidine Gluconate at the Central Venous
Catheter Insertion Site
Subchapter 16.4. Other Practices
Chapter 17. Prevention of Ventilator-Associated Pneumonia (VAP)
Subchapter 17.1. Patient Positioning: Semi-recumbent Positioning and
Continuous Oscillation
Subchapter 17.2. Continuous Aspiration of Subglottic Secretions
Subchapter 17.3. Selective Digestive Tract Decontamination
Subchapter 17.4. Sucralfate and Prevention of VAP
118

Chapter 12. Practices to Improve Handwashing Compliance
Ebbing Lautenbach, MD, MPH, MSCE
University of Pennsylvania School of Medicine
Background
Hospital-acquired infections exact a tremendous toll, resulting in increased morbidity and
mortality, and increased health care costs. 1,2 Since most hospital-acquired pathogens are
transmitted from patient to patient via the hands of health care workers, 3 handwashing is the
simplest and most effective, proven method to reduce the incidence of nosocomial infections. 4
Indeed, over 150 years ago, Ignaz Semmelweis demonstrated that infection-related mortality
could be reduced when health care personnel washed their hands. 5 A recent review summarized
the 7 studies published between 1977 and 1995 that examined the relationship between hand
hygiene and nosocomial infections. 6
Most of the reports analyzed in this study reveal a temporal relation between improved
hand hygiene and reduced infection rates. 6 Despite this well-established relationship, compliance
with handwashing among all types of health care workers remains poor. 7-11 Identifying effective
methods to improve the practice of handwashing would greatly enhance the care of patients and
result in a significant decrease in hospital-acquired infections.
Practice Description
This chapter focuses on practices that increase compliance with handwashing, rather than
the already proven efficacy of handwashing itself. 4 The term “handwashing” defines several
actions designed to decrease hand colonization with transient microbiological flora, achieved
either through standard handwashing or hand disinfection. 4 Standard handwashing refers to the
action of washing hands in water with detergent to remove dirt and loose, transient flora. Hand
disinfection refers to any action where an antiseptic solution is used to clean the hands (ie,
medicated soap or alcohol). Handwashing with bland soap (without disinfectant) is inferior to
handwashing with a disinfecting agent. 12 Hygienic hand rub consists of rubbing hands with a
small quantity (2-3mL) of a highly effective and fast acting antiseptic agent. Because alcohols
have excellent antimicrobial properties and the most rapid action of all antiseptics, they are the
preferred agents for hygienic hand rub (also called waterless hand disinfection). Also, alcohols
dry very rapidly, allowing for faster hand disinfection. 4
Given health care workers’ documented low compliance with recommended
handwashing practices, 7-9 improving compliance represents a more pressing patient safety
concern than does the choice of different disinfectants, or attention to other specific issues such
as choice of drying method, removal of rings, etc. Of the 14 studies reviewed in this chapter
(Table 12.1), all study sites utilized hygienic hand rub and/or another method of hand
disinfection as standard practice. However, only 2 studies assessed the specific characteristics of
handwashing practice (eg, duration of washing, method of drying) according to established
hospital guidelines, 13,14 while the other 12 studies assessed only whether or not handwashing
occurred after patient contact.
119

Prevalence and Severity of the Target Safety Problem
Nosocomial infections occur in about 7-10% of hospitalized patients 1 and account for
approximately 80,000 deaths per year in the United States. 15 Although handwashing has been
proven to be the single most effective method to reduce nosocomial infections, compliance with
recommended hand hygiene practices is unacceptably low. 7-9 Indeed, a recent review of 11
studies noted that the level of compliance with basic handwashing ranged from 16% to 81%. 4 Of
these 11 studies, only 2 noted compliance levels above 50%. 4 One reason for poor handwashing
compliance may be that the importance of this simple protocol for decreasing infections is
routinely underestimated by health care workers. 2 Recent surveys demonstrate that although
most health care workers recognize the importance of handwashing in reducing infections, they
routinely overestimate their own compliance with this procedure. 10 A survey of approximately
200 health care workers noted that 89% recognized handwashing as an important means of
preventing infection. 10 Furthermore, 64% believed they washed their hands as often as their
peers, and only 2% believed that they washed less often than their peers did. 10
Opportunities for Impact
Given these findings, opportunities for improvement in current practice are substantial,
and efforts to improve current practice would have vast applicability. Many risk factors for noncompliance
with hand hygiene guidelines have been identified, including professional category
(eg, physician, nurse, technician), hospital ward, time of day or week, and type and intensity of
patient care. 8 These results suggest that interventions could be particularly targeted to certain
groups of health care workers or to particular locations, to increase the likelihood of compliance.
Importantly, this study demonstrates that the individuals with the highest need for hand hygiene
(ie, those with the greatest workloads) were precisely the same group least likely to wash their
hands. Finally, another recent study noted that approximately 75% of health care workers
surveyed reported that rewards or punishments would not improve handwashing, but 80%
reported that easy access to sinks and availability of hand washing facilities would lead to
increased compliance. 10
Study Designs
A structured search of the PubMed database (including MEDLINE) and review of the
bibliographies of relevant articles identified 14 studies that have examined methods to improve
handwashing compliance (Table 12.1). Three studies were non-randomized controlled trials
(Level 2) that directly compared separate units, or parts of units, in which one area received the
intervention and another did not. 14,16,17 Eleven studies were before-after studies (Level 3), in
which baseline data regarding handwashing rates were obtained during an initial observation
period, and then measured again in the time period after a particular intervention. Regardless of
the type of study design, details regarding the comparability of the groups under observation
were reported in only 4 studies. 13,16,18,19
Study Outcomes
All of the studies reported changes in percent compliance with handwashing, assessing
whether or not handwashing took place (Table 12.1). While 13 studies assessed handwashing
through observation of health care worker behavior (Level 2), one study assessed soap usage as
an indicator of handwashing frequency (Level 3). 20 Two studies also assessed changes in the
quality of handwashing. 13,14 Several studies reported results of surveys conducted following
120

interventions to assess effectiveness and potential adverse events related to the
interventions. 14,20,21 One study also assessed changes in 2 clinical outcomes (incidence of
nosocomial infections and newly detected cases of methicillin-resistant Staphylococcus aureus)
as a result of interventions (Level 1). 18
Evidence for Effectiveness of the Practice
Since many different risk factors have been identified for non-compliance with
handwashing, it is not surprising that a variety of different interventions have been studied in an
effort to improve this practice. While most of the reviewed studies demonstrated significant
improvement in handwashing compliance, 9,13,17,18,20-22 some did not. 14,19,23,24 No single strategy
has consistently been shown to sustain improved compliance with handwashing protocols. 11 In
fact, of the studies which assessed longer-term results following intervention, 16,21,25 all 3 found
that compliance rates decreased from those immediately following the intervention, often
approaching pre-intervention levels.
Potential for Harm
While no harm is likely to befall a patient as a result of handwashing, one potential
adverse effect of handwashing for health care workers is skin irritation. Indeed, skin irritation
constitutes an important barrier to appropriate compliance with handwashing guidelines. 27 Soaps
and detergents can damage the skin when applied on a regular basis. Alcohol-based preparations
are less irritating to the skin, and with the addition of emollients, may be tolerated better. 6
Another potential harm of increasing compliance with handwashing is the amount of time
required to do it adequately. Current recommendations for standard handwashing suggest 15-30
seconds of handwashing is necessary for adequate hand hygiene. 28 Given the many times during
a nursing shift that handwashing should occur, this is a significant time commitment that could
potentially impede the performance of other patient care duties. In fact, lack of time is one of the
most common reasons cited for failure to wash hands. 11 Since alcohol-based handrubs require
much less time, it has been suggested that they might resolve this concern. In fact, a recent study
which modeled compliance time for handwashing as compared with alcoholic rubs, suggested
that, given 100% compliance, handwashing would consume 16 hours of nursing time per
standard day shift, while alcohol rub would consume only 3 hours. 29
Costs and Implementation
Interventions designed to improve handwashing may require significant financial and
human resources. This is true both for multifaceted educational/feedback initiatives, as well as
for interventions that require capital investments in equipment such as more sinks, automated
sinks, or new types of hand hygiene products. The costs incurred by such interventions must be
balanced against the potential gain derived from reduced numbers of nosocomial infections.
Only one study addressed the cost implications of handwashing initiatives. 20 The implementation
of a patient education campaign, when compared to the estimated $5000 per episode cost of each
nosocomial infection, would result in an annual savings of approximately $57,600 for a 300-bed
hospital with 10,000 admissions annually. 20 As others have estimated that the attributable cost of
a single nosocomial bloodstream infection is approximately $40,000 per survivor, 30 the potential
cost savings of interventions to improve handwashing may be even greater.
121

Comment
While many studies have investigated a variety of interventions designed to improve
compliance with handwashing, the results have been mixed. Even when initial improvements in
compliance have been promising, long-term continued compliance has been disappointing.
Future studies should focus on more clearly identifying risk factors for non-compliance, and
designing interventions geared toward sustainability. Some investigators postulate that better
understanding of behavior theory, and its application to infection control practices, might result
in more effectively designed interventions. 26 In addition, any intervention must target reasons for
non-compliance at all levels of health care (ie, individual, group, institution) in order to be
effective. A more detailed study of the cost (and potentially cost savings) of handwashing
initiatives would also foster greater enthusiasm among health care institutions to support such
initiatives.
122

Table 12.1. Fourteen studies of practices to improve handwashing compliance*
Study Setting; Practice Study Design,
Outcomes
All medical staff in a neurologic ICU and a surgical
ICU in a 350-bed tertiary care teaching hospital in
Washington, DC, 1983-84; multifaceted intervention
(education, automatic sinks, feedback) 16
Medical staff in 2 ICUs in a university teach hospital in
Philadelphia; increase number of available sinks 17
Medical staff in a 6-bed post-anesthesia recovery room
and a 15-bed neonatal ICU in a tertiary care hospital in
Baltimore, 1990; automatic sink compared with
standard sink 14
All staff at a large acute-care teaching hospital in
France, 1994-97; hand hygiene campaign including
posters, feedback, and introduction of alcohol-based
solution 18
Medical staff in a 6-bed pediatric ICU in a large
academic medical center in Virginia, 1982-83;
mandatory gowning 19
Medical staff in 2 ICUs in a community teaching
hospital in Tennessee, 1983-84; sequential
interventions of lectures, buttons, observation, and
feedback 24
Medical staff in an 18-bed ICU in a tertiary care
hospital in Australia; introduction of chlorhexidinebased
antiseptic handrub lotion 9
12 nurses in a 12-bed ICU in Mississippi, 1990;
education/feedback intervention 31
Medical staff in an 18-bed pediatric ICU in a children’s
teaching hospital in Melbourne, 1994; 5-step behavioral
modification program 25
Medical staff in a 3000-bed tertiary care center in
France, 1994-95; 13-step handwashing protocol 13
123
Level 2, Level
2
Level 2, Level
2
Level 2, Level
2
Level 3, Level
1
Level 3, Level
2
Level 3, Level
2
Level 3, Level
2
Level 3, Level
2
Level 3, Level
2
Level 3, Level
2
Handwashing Compliance
(unless otherwise noted)†
69% vs. 59% (p=0.005)
76% vs. 51% (p

Table 12.1. Fourteen studies of practices to improve handwashing compliance (cont.)
Medical staff in two ICUs at a teaching hospital in
Virginia, 1997; 6 education/feedback sessions followed
by introduction of alcohol antiseptic agent 22
Medical staff in a 14-bed ICU in a tertiary care
hospital in France, 1998; introduction of alcohol-based
solution 21
All staff in a medical ICU and step-down unit in a large
teaching hospital in Virginia; installation of alcoholbased
solution 23
Medical staff on 2 general inpatient floor at each of 4
community hospitals in New Jersey; patient education
intervention 20
* ICU indicates intensive care unit.
† Results are reported as intervention group vs. control group.
124
Level 3, Level
2
Level 3, Level
2
Level 3, Level
2
Level 3, Level
3
Baseline 22%;
Education/feedback 25%;
Alcohol antiseptic 48%;
(p

References
1. Haley RW, Culver DH, White JW, Morgan WM, Emori TG, Munn VP. The efficacy of
infection surveillance and control programs in preventing nosocomial infection in US
hospitals. Am J Epidemiol. 1985;121:182-205.
2. Jarvis WR. Handwashing-the Semmelweis lesson forgotten? Lancet. 1994;344:1311-2.
3. Larson E. A causal link between handwashing and risk of infection? Examination of the
evidence. Infect Control Hosp Epidemiol. 1988;9:28-36.
4. Pittet D. Improving compliance with hand hygiene in hospitals. Infect Control Hosp
Epidemiol. 2000;21:381-386.
5. Newsom SW. Pioneers in infection control. Ignaz Philipp Semmelweis. J Hosp Infect.
1993;23:175-187.
6. Larson E. Skin hygiene and infection prevention: more of the same or different
approaches? Clin Infect Dis. 1999;29:1287-1294.
7. Doebbeling BN, Stanley GL, Sheetz CT, Pfaller MA, Houston AK, Annis L, et al.
Comparative efficacy of alternative handwashing agents in reducing nosocomial infections
in intensive care units. N Engl J Med. 1992;327:88-93.
8. Pittet D, Mourouga P, Perneger TV. Compliance with handwashing in a teaching hospital.
Ann Intern Med. 1999;130:126-130.
9. Graham M. Frequency and duration of handwashing in an intensive care unit. Am J Infect
Control. 1990;18:77-80.
10. Harris AD, Samore MH, Nafziger R, Dirosario K, Roghmann MC, Carmeli Y. A survey on
handwashing practices and opinions of healthcare workers. J Hosp Infect. 2000;45:318-
321.
11. Larson E, Kretzer EK. Compliance with handwashing and barrier precautions. J Hosp
Infect. 1995;30(Suppl):88-106.
12. Ehrenkranz NJ, Alfonso BC. Failure of bland soap handwash to prevent hand transfer of
patient bacteria to urethral catheters. Infect Control Hosp Epidemiol. 1991;12:654-662.
13. Coignard B, Grandbastien B, Berrouane Y, Krembel C, Queverue M, Salomez JL, et al.
Handwashing quality: impact of a special program. Infect Control Hosp Epidemiol.
1998;19:510-513.
14. Larson E, McGeer A, Quraishi A, Krenzischek D, Parsons BJ, Holdford J, et al. Effect of
an automated sink on handwashing practices and attitudes in high-risk units. Infect Control
Hosp Epidemiol. 1991;12:422-428.
15. Jarvis WR. Selected aspects of the socioeconomic impact of nosocomial infections:
morbidity, mortality, cost, and prevention. Infect Control Hosp Epidemiol. 1996;17:552-
557.
16. Larson EL, Bryan JL, Adler LM, Blane C. A multifaceted approach to changing
handwashing behavior. Am J Infect Control. 1997;25:3-10.
17. Kaplan LM, McGuckin M. Increasing handwashing compliance with more accessible
sinks. Infect Control Hosp Epidemiol. 1986;7:408-410.
18. Pittet D, Hugonnet S, Harbarth S, Mourouga P, Sauvan V, Touveneau S, et al.
Effectiveness of a hospital-wide programme to improve compliance with hand hygiene.
Lancet. 2000;356:1307-1312.
19. Donowitz LG. Handwashing technique in a pediatric intensive care unit. Am J Dis Child.
1987;141:683-685.
125

20. McGuckin M, Waterman R, Porten L, Bello S, Caruso M, Juzaitis B, et al. Patient
education model for increasing handwashing compliance. Am J Infect Control.
1999;27:309-314.
21. Maury E, Alzieu M, Baudel JL, Haram N, Barbut F, Guidet B, et al. Availability of an
alcohol solution can improve hand disinfection compliance in an intensive care unit. Am J
Respir Crit Care Med. 2000;162:324-327.
22. Bischoff WE, Reynolds TM, Sessler CN, Edmond MB, Wenzel RP. Handwashing
compliance by health care workers: the impact of introducing an accessible, alcohol-based
hand antiseptic. Arch Intern Med. 2000;160:1017-1021.
23. Muto CA, Sistrom MG, Farr BM. Hand hygiene rates unaffected by installation of
dispensers of a rapidly acting hand antiseptic. Am J Infect Control. 2000;28:273-276.
24. Simmons B, Bryant J, Km N, Spencer L, Arheart K. The role of handwashing in prevention
of endemic intensive care unit infections. Infect Control Hosp Epidemiol. 1990;11:589-
594.
25. Tibballs J. Teaching hospital medical staff to handwash. Med J Aust. 1996; 164: 395-398.
26. Kretzer EK, Larson EL. Behavioral interventions to improve infection control practices.
Am J Infect Control. 1998;26:245-253.
27. Larson E. Handwashing and skin: physiologic and bacteriologic aspects. Infect Control.
1985;6:14-23.
28. Larson EL. APIC Guidelines for handwashing and hand antisepsis in health care settings.
Am J Infect Control. 1995;23:251-269.
29. Voss A, Widmer AF. No time for handwashing!? Handwashing versus alcoholic rub: can
we afford 100% compliance? Infect Control Hosp Epidemiol. 1997;18:205-208.
30. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream infection in critically ill patients.
Excess length of stay, extra costs, and attributable mortality. JAMA. 1994;271:1598-1601.
31. Dubbert PM, Dolce J, Richter W, Miller M, Chapman SW. Increasing ICU staff
handwashing: effects of education and group feedback. Infect Control Hosp Epidemiol.
1990;11:191-193.
126

Chapter 13. Impact of Barrier Precautions in Reducing the Transmission of
Serious Nosocomial Infections
Ebbing Lautenbach, MD, MPH, MSCE
University of Pennsylvania School of Medicine
Background
Many nosocomial infections are easily transferable from patient-to-patient, either via the
hands of health care workers, 1,2 or through the contamination of inanimate objects, including
clothing and equipment. 3,4 For some infections, the threat to other patients is considered serious
enough that many institutions employ special barrier precautions, such as the use of gloves,
gowns and disposable equipment for all patient contact, in caring for patients colonized or
infected with these pathogens. Vancomycin-resistant enterococci (VRE) 5 and Clostridium
difficile 6 represent 2 typical examples of nosocomial pathogens that may trigger such
precautions.
Although adherence to barrier precautions to prevent the spread of particularly
concerning nosocomial pathogens has obvious face validity, the utility of specific interventions
and the optimal forms they should take remain unclear. This uncertainty may in part reflect the
impact of particular aspects of the epidemiology of the targeted nosocomial pathogens – ie, the
benefit of a given strategy may vary in different settings and with different organisms.
Consequently, this chapter contrasts with the review of handwashing (Chapter 13), a practice for
which the benefit was regarded as sufficiently established to warrant focusing on strategies for
improving compliance. While compliance with barrier precautions is also an important topic and
likely plays a significant role in the efficacy of such interventions, this chapter analyzes the
literature evaluating the benefit of the barrier precautions themselves.
Practice Description
Barrier precautions include any activity designed to prevent the spread of nosocomial
pathogens from patient to patient. This chapter reviews the following 3 practices:
• Use of gowns and gloves for all contact with patients colonized or infected
with VRE and/or C. difficile: Health care workers typically don gloves and
gowns when entering the room of an infected or colonized patient, and
remove them upon exiting (followed immediately by handwashing) to reduce
the likelihood of clothing or equipment contamination that could transmit
pathogens to other patients;
• Use of dedicated or disposable examining equipment for patients colonized or
infected with VRE and/or C. difficile: Hospital equipment (ie, blood pressure
cuffs, thermometers) remains in a patient’s room and is not carried from room
to room; and
• Patient and/or staff cohorting for patients colonized or infected with VRE
and/or C. difficile: Patients colonized or infected with similar pathogens are
admitted to specific floors of the hospital where designated health care
workers care only for patients colonized or infected with these pathogens.
127

Prevalence and Severity of the Target Safety Problem
Nosocomial infections, including C. difficile-associated diarrhea and VRE, significantly
increase the morbidity and mortality of hospitalized patients. 5,6 Both infections are also
associated with increased hospital costs. Recent evidence also suggests there may be a
relationship between C. difficile and VRE, with C. difficile infection identified as a risk factor for
VRE infection. 7 The increased incidence of both VRE and C. difficile can be attributed to spread
from patient to patient. 5,6 Failure to recognize these dissemination patterns may result in an
inability to contain outbreaks when they occur in the hospital.
C. difficile has been identified as the major, if not only, important cause of infectious
diarrhea that develops in patients after hospitalization, occurring in up to 30% of adult
hospitalized patients who developed diarrhea. 5 One study found an acquisition rate of 13% for
patients hospitalized 1-2 weeks, which increased to 50% for patients hospitalized >4 weeks. 8 In
addition, the incidence of C. difficile infection has increased in recent years, with one study
reporting a 5-fold increase in clinical infection between 1993 and 1996. 9 C. difficile infection
increases lengths of stay, often to as long as 18-30 days 10,11 and, when fulminant, can lead to
exploratory and therapeutic surgical procedures. 12 Mortality attributable to C. difficile, while
reported, occurs in fewer than 5% of patients. 13 The costs associated with C. difficile diarrhea,
while not well described, may be as high as $10,000 per patient. 14
VRE, first described in 1988, currently accounts for greater than 25% of all nosocomial
enterococci. 6 Early national data suggested that infections with VRE were associated with
mortality rates of over 36%, more than double that of patients with vancomycin-susceptible
(VSE) infections. 15 While later studies called some of these results into question, 16,17 the most
recent studies have again suggested that vancomycin-resistance carries an independent effect on
mortality. 18 VRE infections are also associated with significantly higher hospital costs than those
due to VSE. 18
Although C. difficile and VRE are among the most common nosocomial pathogens that
have significant effects on morbidity, mortality, and cost, there are a number of other nosocomial
pathogens which could also be studied. These include pathogens such as methicillin-resistant
Staphylococcus aureus (MRSA), extended-spectrum beta-lactamase (ESBL) producing
Enterobacteriaceae, Acinetobacter species, and Pseudomonas aeruginosa. While these are all
important nosocomial pathogens, C. difficile and VRE were chosen as examples because they are
extremely common, and they represent both antibiotic-susceptible (C. difficile) and antibioticresistant
(VRE) pathogens. Additionally, (unlike MRSA and P. aeruginosa) the epidemiology of
both pathogens is complex, representing both person-to-person spread and association with prior
antibiotic use, allowing for a more comprehensive discussion of the relative merits of both
antimicrobial use interventions and barrier precaution interventions (see Chapter 15 for more
discussion regarding other antimicrobial intervention practices) and their general application to
other pathogens.
Opportunities for Impact
As noted above, both VRE and C. difficile affect a large proportion of hospitalized
patients. Improvements in barrier precaution interventions against these pathogens would have a
tremendous impact. There are few data regarding the percentage of hospitals that employ any
one of a number of barrier precautions (eg, gowns, gloves, disposable thermometers). 19 In
addition, while standard practice is to apply barrier precautions for patients with nosocomial
pathogens with demonstrated horizontal spread, compliance with precautions is frequently
128

poor, 20 often below 50%. 21 Purported reasons for this lack of compliance include lack of
resources and busy staff workload. 20 Regardless, these results suggest that the opportunity for
improvement in these practices is great.
Study Designs
A structured search of the PubMed database (including MEDLINE) and review of the
bibliographies of relevant articles identified 19 studies that have examined the implementation of
barrier precaution practices designed to impact the incidence of VRE and/or C. difficile infection
(Table 13.1, 13.2, 13.3). All studies found on literature search were included in this review
except for those reporting very small outbreaks (defined as fewer than 10 cases of C. difficile or
VRE). Sixteen of the reviewed studies were before-after observational cohort studies (Level 3),
in which baseline data regarding incidence of VRE or C. difficile were obtained during an
observational period and compared to a second period after implementation of an intervention.
Crude comparability data on the before and after groups (eg, total admissions, patient census)
were provided in 2 reports 22,23 while only one study statistically compared the before and after
groups to assess comparability. 24 Three reports 25-27 detailed unblinded comparative studies
(Level 2) in which patients on different wards were assigned different interventions. Each of
these studies assessed the comparability of the study groups on the basis of underlying
demographic variables.
Study Outcomes
All of the studies reviewed reported changes in the incidence or prevalence of either VRE
or C. difficile as a result of barrier precaution interventions (Level 1). For studies investigating C.
difficile, all outcomes were reported in terms of clinical infections. For studies investigating
VRE, outcomes were reported as VRE colonization and/or infection rates.
Evidence for Effectiveness of the Practice
As both VRE and C. difficile have clearly been shown to be transferable from patient-topatient,
interventions designed to improve barrier precautions yield significant reductions in the
incidence of infection with these two pathogens. All studies that examined the effect of enhanced
barrier precautions on C. difficile infection demonstrated benefit, suggesting that barrier
precaution interventions are effective in controlling its emergence. Most studies employed a
multifaceted approach including several different barrier precaution components. For example,
one study combined use of vinyl gloves and ongoing educational interventions, 26 another
included cohorting, culture screening, and daily room disinfection, 28 while another combined
reinforcement of enteric precautions, replacement of electronic thermometers, and institution of
closed paper towel dispensers. 29 Given the varied components of barrier precaution interventions
instituted in different studies, it is difficult to determine the specific effect of any individual
component.
The evidence of effectiveness of barrier precautions for VRE is somewhat less clear-cut.
All but 4 27,30-32 of the studies examining the effect of barrier precautions on VRE demonstrated a
benefit, but study design differences and particular epidemiologic trends may account for the
inconsistent findings.
One of the 4 studies that noted no significant effect compared glove use to glove and
gown use. 27 The second 30 noted that the emergence of VRE at the study institution was due to
multiple genetically-unrelated strains, suggesting that person-to-person spread was less
important at that site. It is thus not surprising that barrier precautions would have less of an
129

effect. In the third study, 32 routine rectal swab surveillance and contact precautions were
instituted in response to a clinical outbreak of VRE and surveillance was continued for only 6
months. Since surveillance was not conducted prior to institution of precautions, it is impossible
to say what the colonization prevalence had been prior to the intervention. Furthermore, as the
authors point out, it may be that the outbreak would have been much worse had the precautions
not been put in place. Finally, no determination of genetic relatedness (and hence spread) was
made in this study. In the fourth study, 31 while there was a reduction in the isolation of VRE,
there was not complete eradication. According to the authors, the most likely reason for this lessthan-optimal
response was poor compliance with contact precaution guidelines.
Thus, it appears that enhanced barrier precautions are generally effective in reducing the
incidence of VRE but that various aspects of both the epidemiology of the VRE outbreak and the
implementation of guidelines may temper the effectiveness of interventions. Similar to the
studies investigating response of C. difficile to barrier precautions, most studies of VRE
employed several components of barrier precautions as part of a multifaceted approach (Table
13.1). It is thus difficult to determine the specific effect of any individual component.
Potential for Harm
None of the reviewed studies reported any assessment of possible harm as a result of the
barrier precaution interventions. In fact, the implementation of barrier precautions is unlikely to
result in harm to the patient. One potential concern is that time necessary to comply with the
interventions (eg, gowning, gloving), might make health care workers less likely to complete
tasks necessary to provide acceptable patient care. Indeed, it has recently been noted that health
care workers were half as likely to enter the rooms of patients on contact isolation. 33
Furthermore, while contact precautions appeared to have little effect on patient examination by
resident physicians, attending physicians were 50% less likely to examine a patient on contact
precautions compared to a patient not on precautions. 34 Future studies should address these
concerns by documenting the time required to adhere to barrier precautions, and determining the
potential impact of precautions on patient care.
Another potentially harmful consequence of barrier precaution interventions is the
psychological effect that contact precautions may have on the isolated patient. While research
has examined the effects of sensory deprivation and social isolation, a recent review of the
literature noted little progress in the investigation of the psychological effects of contact
isolation. 35
Costs and Implementation
It seems apparent that the more complicated an intervention, the less likely health care
workers will adhere to it. While 2 studies noted compliance with barrier precautions at close to
90%, 21,24 others noted levels closer to 70%. 31 One study actually noted compliance levels to be
significantly higher in those health care workers who used both gowns and gloves compared to
those using only gowns. 27 This somewhat counterintuitive finding suggests that other factors
may be at play in influencing compliance. Of the reviewed studies that reported compliance
levels, all did so relatively shortly after the initial implementation of interventions. Future studies
should assess compliance with guidelines over a longer period.
Four studies reported the costs of specific interventions. Implementation of use of
disposable thermometers was estimated at $14,055 per year at a 343-bed institution. 22 Another
study of the impact of disposable thermometers estimated that the cost per prevented C. difficile
infection would be approximately $611. 25 A study using a multifaceted approach estimated that
130

the annual expenses due directly to increased demand for gowns and gloves were approximately
$11,000. 31 Finally, a multifaceted intervention at a 254-bed long-term care facility which
included education, gowns and gloves for resident contact, no sharing of personal equipment,
and daily double cleaning of resident rooms and wheelchairs, estimated the total cost of the
intervention to be $12,061 Canadian (approximately $8000 US). 36
The costs of implementing a program to enhance barrier precaution practices must be
balanced against the potential cost savings due to decreased incidence of nosocomial infections.
Both VRE and C. difficile infections have been associated with significantly increased length of
hospital stay. 5,6 Preventing even a small number of these infections is likely to have a significant
financial impact. While several of the reviewed studies documented costs associated with
various interventions, 22,25,26,31,36 no study systematically compared these costs to the potential
cost savings of infections prevented.
Comment
The majority of reviewed studies demonstrated a significant reduction in the incidence of
VRE or C. difficile following barrier precaution interventions. The fact that not all studies found
a benefit suggests that future studies should identify those scenarios (eg, outbreak, endemic
colonization, etc.) in which attention to barrier precautions is most likely to be beneficial. In
addition, it is possible that a combined intervention involving both enhanced barrier precautions
as well as antibiotic formulary interventions might be needed in order to effect the greatest
possible change in VRE and C. difficile infection rates. While these studies, much like those that
examined the impact of antibiotic use practices, demonstrated short-term success, future studies
should determine the efficacy of such interventions over the long term. Finally, the costeffectiveness
of such strategies should be investigated.
131

Table 13.1. Studies of multifaceted approaches with and without “cohorting”*
Study Setting Compliance Study
Design,
Outcomes
725-bed academic medical center in
Philadelphia in 1987-88: beforeafter
study of impact of
multifaceted intervention (isolation
precautions, clindamycin
restriction) on C. difficile 37
350-bed acute care hospital in
Virginia in 1987-96: before-after
study of impact of multifaceted
intervention on C. difficile
infections 23
840-bed tertiary care center in
Brussels in 1989-90: impact of a
multifaceted infection control
intervention, including cohorting,
on incidence of C. difficile 28
Bone marrow transplant unit of a
large academic medical center in
Texas in 1995: impact of
multifaceted infection control
intervention on C. difficile attack
rate 29
Tertiary-care Veterans Affairs
Medical Center in Brooklyn in
1991-95: impact of multifaceted
infection control intervention on
VRE rates 30
22-bed oncology unit in a 650-bed
tertiary care hospital in New York
in 1993-95: impact of multifaceted
infection control program, including
cohorting, on VRE infection and
colonization 24
375-bed community hospital in
Indianapolis in 1995-96: impact of
cohorting on VRE prevalence 21
NA Level 3,
Level 1
NA Level 3,
Level 1
NA Level 3,
Level 1
NA Level 3,
Level 1
NA Level 3,
Level 1
91.7% of
persons who
entered room
used gowns
and gloves
appropriately
Compliance
with recommendations
rose from
22% to 88%
(p

254-bed long-term care facility in
Toronto in 1996-97: impact of
barrier precautions including
cohorting on prevalence of VRE 36
NA Level 3,
Level 1
133
4/85 (4.7%) patients initially
screened were VRE colonized.
No patients in subsequent
screenings were positive.

Table 13.1. Studies of multifaceted approaches with and without “cohorting” (cont.)
23-bed oncology unit in a 1300bed
teaching hospital in South
Africa in 1998: impact of barrier
precautions including cohorting
on VRE prevalence 39
347-bed tertiary care medical
center in Massachusetts in 1993:
impact of a multifaceted infection
control intervention including
cohorting on VRE infection and
colonization 31
NA Level 3,
Level 1
Overall
handwashing
complianc
e was 71%
134
Level 3,
Level 1
* NA indicates not applicable; VRE, vancomycin-resistant enterococci.
VRE colonization decreased from
19/34 (55%) patients to 1/14 (7%)
following implementation of
infection control interventions
In the year prior interventions,
116 patients were colonized or
infected with VRE, compared
with 126 in the year after
implementation.

Table 13.2. Studies of barrier precaution interventions*
Study Setting Compliance Study
Design,
Outcomes
370-bed academic medical center in
Massachusetts in 1991-92: beforeafter
study of impact of infection
control interventions on C. difficile
incidence 38
Veterans Administration Medical
Center in Minnesota in 1986-87:
impact of universal glove use on
incidence of C. difficile 26
8-bed combined medical and surgical
ICU in a 235-bed acute care hospital
in New York City in 1990-91: impact
of barrier precautions on VRE
colonization 1
250-bed university-affiliated hospital
in Rhode Island in 1991-92: impact
of sequential barrier precaution
intervention on VRE 40
181 consecutive patients admitted to
the medical ICU in a 900-bed urban
teaching hospital in Chicago in 1994-
95: comparison of impact of gown
and glove vs. glove on incidence of
VRE colonization 27
550-bed tertiary teaching hospital in
Minneapolis in 1993-94: impact of
barrier precautions on VRE
colonization 32
NA Level 3,
Level 1
Mean glove
use/100 pts:
4539 on glove
ward; 3603 on
control ward
(p=NS)
135
Level 2,
Level 1
NA Level 3,
Level 1
NA Level 3,
Level 1
Compliance in
glove and
gown group,
79%; glove
group, 62%
(p

Table 13.3. Studies of use of dedicated or disposable examining equipment*
Study Setting Complianc
e
343-bed acute hospital and
538-bed skilled nursing
facility in New York: beforeafter
study of impact of
replacing electronic
thermometers with disposable
thermometers on C. difficile
infection rate 22
20 inpatient units in a 700-bed
university hospital in
Virginia: randomized
crossover trial of impact of
disposable thermometers for
prevention of C. difficile 25
343-bed acute care facility in
New York in 1992: impact of
change to tympanic
thermometers on VRE
incidence 22
100%
replacemen
t of
electronic
thermometers
100%
compliance
with use of
specific
types of
thermometers
100%
switch to
tympanic
thermometers
Study
Design,
Outcomes
Level 3,
Level 1
Level 2,
Level 1
Level 3,
Level 1
136
Change in C. difficile or VRE
Incidence of C. difficile decreased
from 2.71 to 1.76 cases per 1000
patients in the acute hospital
(p

References
1. Handwerger S, Raucher B, Altarac D, Monka J, Marchione S, Singh KV, et al. Nosocomial
outbreak due to Enterococcus faecium highly resistant to vancomycin, penicillin, and
gentamicin. Clin Infect Dis. 1993;16:750-755.
2. Chang VT, Nelson K. The role of physical proximity in nosocomial diarrhea. Clin Infect
Dis. 2000;31:717-722.
3. Byers KE, Durbin LJ, Simonton BM, Anglim AM, Adal KA, Farr BM. Disinfection of
hospital rooms contaminated with vancomycin-resistant Enterococcus faecium. Infect
Control Hosp Epidemiol. 1998;19:261-264.
4. Mayfield JL, Leet T, Miller J, Mundy LM. Environmental control to reduce transmission of
Clostridium difficile. Clin Infect Dis. 2000;31:995-1000.
5. Johnson S, Gerding DN. Clostridium difficile-associated diarrhea. Clin Infect Dis.
1998;26:1027-1036.
6. Murray BE. Vancomycin-resistant enterococcal infections. N Engl J Med. 2000;342:710-
721.
7. Roghmann MC, McCarter RJ, Brewrink J, Cross AS, Morris JG. Clostridium difficile
infection is a risk factor for bacteremia due to vancomycin-resistant enterococci (VRE) in
VRE-colonized patients with acute leukemia. Clin Infect Dis. 1997;25:1056-1059.
8. Clabots CR, Johnson S, Olson MM, Peterson LR, Gerding DN. Acquisition of Clostridium
difficile by hospitalized patients: evidence of colonized new admissions as the source of
infection. J Infect Dis. 1992;166:561-567.
9. Wilcox MH, Smyth ETM. Incidence and impact of Clostridium difficile infection in the
UK, 1993-1996. J Hosp Infect. 1998;39:181-187.
10. MacGowan AP, Brown I, Feeney R, Lovering A, McCulloch SY, Reeves DS. Clostridium
difficile associated diarrhea and length of hospital stay. J Hosp Infect. 1995;31:241-244.
11. Riley TV, Codde JP, Rouse IL. Increase length of stay due to Clostridium difficile
associated diarrhoea. Lancet. 1995;345:455-456.
12. Kent KC, Rubin MS, Wroblewski L, Hanff PA, Sline W. The impact of Clostridium
difficile on a surgical service. Ann Surg. 1998;227:296-301.
13. Olson MM, Shanholtzer CJ, Lee JT, Gerding DN. Ten years of prospective Clostridium
difficile-associated disease surveillance and treatment at the Minneapolis VA Medical
Center, 1982-1991. Infect Control Hosp Epidemiol. 1994;15:371-381.
14. Spencer RC. Clinical impact and associated costs of Clostridium difficile-associated
disease. J Antimicrob Chemother. 1998;41(Suppl C):C5-12.
15. Centers for Disease Control and Prevention (CDC). Nosocomial enterococci resistant to
vancomycin - United States, 1989-93. MMWR Morb Mortal Wkly Rep. 1993;42:579-597.
16. Lautenbach E, Bilker WB, Brennan PJ. Enterococcal bacteremia: risk factors for
vancomycin resistance and predictors of mortality. Infect Control Hosp Epidemiol.
1999;20:318-323.
17. Linden PK, Pasculle AW, Manez R, Kramer DJ, Fung JJ, Pinna AD, et al. Differences in
outcomes for patients with bacteremia due to vancomycin-resistant Enterococcus faecium
or vancomycin-susceptible E. faecium. Clin Infect Dis. 1996;22:663-670.
18. Stosor V, Peterson LR, Postelnick M, Noskin GA. Enterococcus faecium bacteremia: does
vancomycin resistance make a difference? Arch Intern Med. 1998;158:522-527.
19. Scheckler WE, Brimhall D, Buck AS, Farr BM, Friedman C, Garibaldi RA, et al.
Requirements for infrastructure and essential activities of infection control and
137

epidemiology in hospitals: a consensus panel report. Infect Control Hosp Epidemiol.
1998;19:114-124.
20. Kollef MH, Fraser VJ. Antibiotic resistance in the intensive care unit. Ann Intern Med.
2001;134:298-314.
21. Jochimsen EM, Fish L, Manning K, Young S, Singer DA, Baker R, et al. Control of
vancomycin-resistant enterococci at a community hospital: efficacy of patient and staff
cohorting. Infect Control Hosp Epidemiol. 1999;20:106-109.
22. Brooks SE, Veal RO, Kramer M, Dore L, Schupf N, Adachi M. Reduction in the incidence
of Clostridium difficile-associated diarrhea in an acute care hospital and a skilled nursing
facility following replacement of electronic thermometers with single-use disposables.
Infect Control Hosp Epidemiol. 1992;13:98-103.
23. Zafar AB, Gaydos LA, Furlong WB, Nguyen MH, Mennonna PA. Effectiveness of
infection control program in controlling nosocomial Clostridium difficile. Am J Infect
Control. 1998;26:588-593.
24. Montecalvo MA, Jarvis WR, Uman J, Shar DK, Petrullo C, Rodney K, et al. Infectioncontrol
measures reduce transmission of vancomycin-resistant enterococci in an endemic
setting. Ann Intern Med. 1999;131:269-272.
25. Jernigan JA, Siegman-Igra Y, Guerrant RC, Farr BM. A randomized crossover study of
disposable thermometers for prevention of Clostridium difficile and other nosocomial
infections. Infect Control Hosp Epidemiol. 1998;19:494-499.
26. Johnson S, Gerding DN, Olson MM, Weiler MD, Hughes RA, Clabots CR, et al.
Prospective, controlled study of vinyl glove use to interrupt Clostridium difficile
nosocomial transmission. Am J Med. 1990;88:137-140.
27. Slaughter S, Hayden MK, Nathan C, Hu TC, Rice T, Van Voorhis J, et al. A comparison of
the effect of universal use of gloves and gowns with that of glove use alone on acquisition
of vancomycin-resistant enterococci in a medical intensive care unit. Ann Intern Med.
1996;125:448-456.
28. Struelens MJ, Maas A, Nonhoff C, Deplano A, Rost F, Serruys E, et al. Control of
nosocomial transmission of Clostridium difficile based on sporadic case surveillance. Am J
Med. 1991;91(Suppl 3B):138S-44S.
29. Hanna H, Raad I, Gonzalez V, Umphrey J, Tarrand J, Neumann J, et al. Control of
nosocomial Clostridium difficile transmission in bone marrow transplant patients. Infect
Control Hosp Epidemiol. 2000;21:226-8.
30. Quale J, Landman D, Atwood E, Kreiswirth B, Willey BM, Ditore V, et al. Experience
with a hospital-wide outbreak of vancomycin-resistant enterococci. Am J Infect Control.
1996;24:372-379.
31. Lai KK, Kelley AL, Melvin ZS, Belliveau PP, Fontecchio SA. Failure to eradicate
vancomycin-resistant enterococci in a university hospital and the cost of barrier
precautions. Infect Control Hosp Epidemiol. 1998;19:647-652.
32. Wells CL, Juni BA, Cameron SB, Mason KR, Dunn DL, Ferrieri P, et al. Stool carriage,
clinical isolation, and mortality during an outbreak of vancomycin-resistant enterococci in
hospitalized medical and/or surgical patients. Clin Infect Dis. 1995;21:45-50.
33. Kirkland KB, Weinstein JM. Adverse effects of contact isolation. Lancet. 1999;354:1177-
1178.
34. Higgins LA, Saint S, Nallamothu BK, Chenoweth C. Do physicians examine patients under
contact precautions less frequently? Paper presented at: 24th Annual Meeting of the
Society for General Internal Medicine; May 2-5, 2001; San Diego, CA.
138

35. Gammon J. The psychological consequences of source isolation: a review of the literature.
J Clin Nurs. 1999;8:13-21.
36. Armstrong-Evans M, Litt M, McArthur MA, Willey B, Cann D, Liska S, et al. Control of
transmission of vancomycin-resistant Enterococcus faecium in a long-term care facility.
Infect Control Hosp Epidemiol. 1999;20:312-317.
37. Brown E, Talbot GH, Axelrod P, Provencher M, Hoegg C. Risk factors for Clostridium
difficile toxin-associated diarrhea. Infect Control Hosp Epidemiol. 1990;11:283-290.
38. Lai KK, Melvin ZS, Menard MJ, Kotilainen HR, Baker S. Clostridium difficile-associated
diarrhea: epidemiology, risk factors, and infection control. Infect Control Hosp Epidemiol.
1997;18:628-632.
39. McCarthy KM, Van Nierop W, Duse A, Von Gottberg A, Kassel M, Perovic O, et al.
Control of an outbreak of vancomycin-resistant Enterococcus faecium in an oncology ward
in South Africa: effective use of limited resources. J Hosp Infect. 2000;44:294-300.
40. Boyce JM, Opal SM, Chow JW, Zervos MJ, Potter-Bynoe G, Sherman CB, et al. Outbreak
of multidrug-resistant Enterococcus faecium with transferable vanB class vancomycin
resistance. J Clin Microbiol. 1994;32:1148-53.
139

140

Chapter 14. Impact of Changes in Antibiotic Use Practices on Nosocomial
Infections and Antimicrobial Resistance – Clostridium Difficile and
Vancomycin-resistant Enterococcus (VRE)
Ebbing Lautenbach, MD, MPH, MSCE
University of Pennsylvania School of Medicine
Background
As discussed in the chapters on handwashing and barrier precautions (Chapters 12 and
13), hospital infection control has historically focused on preventing the transmission of
nosocomial pathogens—either from patient to patient or from provider to patient. The potential
role of overseeing hospital-wide antibiotic use as an infection control measure has also been
recognized for many years. 1 With the widespread emergence of nosocomial antibiotic-resistant
infections over the past 10-15 years, institutional efforts to control antibiotic use have become a
priority for infection control. 2,3
The practices reviewed in this chapter involve institutional efforts to control antibiotic
use as a means of controlling complications of antibiotic overuse or misuse. In evaluating the
potential benefits of these practices, the focus is on the impacts of antibiotic use on infections
with vancomycin-resistant enterococci (VRE) 4 and Clostridium difficile. 5 These pathogens
represent two of the most important nosocomial pathogens with relationships to inappropriate
antibiotic use. Moreover, as suggested by recent evidence, infection with C. difficile may
represent a risk factor for infection with VRE. 6
Practice description
Interventions designed to limit the use of antibiotics may take many forms. Specific
practices reviewed in the chapter include:
• Infectious diseases physician approval 7 – all requests for an antibiotic are
discussed with an infectious diseases physician who decides whether use is
appropriate
• Monitoring of antibiotic use by pharmacy service 8 – pharmacists monitor the
use of certain antibiotics and make recommendations for changes to the
prescriber
• Guidelines for antimicrobial use 8 – dissemination to physicians of guidelines
describing appropriate and inappropriate use
• Therapeutic substitution 9 – use of one agent replaced by another agent with
similar spectrum of activity
• Computer-assisted prescribing 10 – computer-based restriction of agent with
extra prompts requesting documentation of indication for agent
• Antibiotic-management program (AMP) 11 – continuation of antibiotic after a
specific duration requires approval from either an infectious diseases
physician or pharmacist on the AMP
141

Prevalence and Severity of the Target Safety Problem
This chapter focuses on 2 of the most important nosocomial pathogens: VRE and C.
difficile. VRE currently accounts for greater than 25% of all nosocomial enterococci 4 and confers
an increased risk of death, independent of comorbid conditions that may have initially led to the
infection. 12 VRE infections are also associated with significantly higher hospital costs than those
due to vancomycin-sensitive enterococci (VSE) 12 (see Chapter 13). C. difficile represents the
major, if not only, important infectious cause of nosocomial diarrhea. 5 Although death
attributable to C. difficile occurs in less than 5% of patients, 17 the impact of C. difficile infection
remains significant. Patients may require substantially longer lengths of hospital stay—upwards
of 18-30 days, 18,19 with exploratory and therapeutic surgical procedures required in severe
cases. 20 It has also been suggested that more debilitated patients (eg, in rehabilitation centers or
long-term care facilities) may be at even greater risk for increased morbidity and mortality due to
C. difficile infection. 21 The costs associated with C. difficile diarrhea, while not well described,
are estimated to be as high as $10,000 per patient 22 (see Chapter 13).
Opportunities for Impact
Over half of all hospitalized patients are treated with antibiotics. 23 The antibiotics
represent a significant portion of overall health care costs, accounting for between 20% and 50%
of total hospital drug expenditures. 23 It has been estimated that 50% of all antibiotics prescribed
are either at the wrong dose, the wrong drug, or taken for the wrong duration. 24,25 These findings
suggest that there is significant room for improvement in antibiotic prescribing practices.
Most hospitals employ formulary restrictions for certain medications (particularly
expensive agents, selecting one drug from a group of equivalent agents). However, only a
minority of hospitals uses formulary restrictions to limit the use of entire antibiotic classes or
specific agents. Those hospitals that do employ antimicrobial formulary restrictions most often
do so as a means of controlling costs, rather than as an infection control measure. 26 Thus, there
remain substantial opportunities to expand upon these existing formulary programs to control the
emergence of antimicrobial resistance.
Study Designs
A structured search of the PubMed database (including MEDLINE) and review of the
bibliographies of relevant articles identified 10 studies that have examined methods to change
antibiotic use with respect to VRE and/or C. difficile infection (Table 14.1). All of these studies
were before-after observational cohort studies (Level 3) in which baseline data regarding
incidence of VRE or C. difficile were obtained during an observational period and compared
with a second time period after an intervention had been implemented. Data on baseline
comparability of the before and after groups were not reported in 6 studies. 8-11,21,27 Two studies
only reported similar admission and census rates during the before and after time periods, 7,28
while 2 studies compared patients in the 2 time periods on the basis of numerous variables. 29,30
Study Outcomes
All of the studies reviewed reported changes in the clinical incidence or prevalence of
either VRE or C. difficile as a result of antibiotic practice interventions (Level 1). Studies
investigating C. difficile measured clinical infections. Studies investigating VRE examined VRE
infection 11 or VRE colonization. 8-10,27
142

Evidence for Effectiveness of the Practice
Of the 10 studies listed in Table 14.1, all but 3 8,11,21 showed significant reductions in the
incidence of C. difficile or VRE following practice changes. Several possibilities may explain the
negative findings of these 3 studies. First, the interventions analyzed might not have produced
significant alterations in antibiotic use, so that infection rates with the target pathogens remained
unchanged. Second, it is possible that patient-to-patient spread of these pathogens limited the
efficacy of the interventions, as this mode of transmission is well known to occur for both VRE
and C. difficile, usually via the hands of health care workers (see also Chapter 13). 31,32 Third,
since environmental contamination occurs with both these pathogens, successful control of these
organisms may require enhanced disinfection procedures in some cases. 33,34 Targeting antibiotic
use may not be sufficient to reduce incidence of these pathogens since a significant number of
infected or colonized patients may serve as reservoirs. Under this scenario, the argument for
barrier precautions as an adjunct measure to prevent spread of organisms from patient to patient
becomes more persuasive (see Chapter 13). Indeed, although changes in antibiotic use practice
were the primary intervention in all of the studies reviewed here, one study included a
component of enhanced barrier precautions in the intervention. 21 Future studies should
investigate the impact of such multifaceted interventions, both for VRE and C. difficile as well as
for other nosocomial pathogens.
Other Potential Benefits
Although not the focus of this chapter, the practices reviewed here may have a beneficial
impact on other emerging nosocomial pathogens strongly associated with inappropriate
antibiotic use, such as extended-spectrum beta-lactamase (ESBL) producing
Enterobacteriaceae. 35 In addition, although we have focused on control of VRE as an end in
itself, a primary motivation to achieve this goal is the need to delay the emergence of
vancomycin-resistance in Staphylococcus aureus. 36,37 As S. aureus represents the most common
nosocomial infection, 38 the development of high-level vancomycin resistance among
staphylococci would constitute a public health disaster. 39
Thus, practices that decrease the prevalence of VRE may play an important, albeit
indirect, role in preventing or delaying this occurrence.
Potential for Harm
Few of the reviewed studies reported any assessment of possible harm as a result of the
antibiotic use practice interventions. One potential result of interventions designed to reduce the
use of one antibiotic, or antibiotic class, is the subsequent increase in the use of another class of
agents to compensate. In fact, one reviewed study 7 noted an increase in the use of other antianaerobic
agents as clindamycin use decreased. Whether changes in antibiotic use results in
changes in antimicrobial susceptibilities, either in the pathogen under study (eg, VRE, C.
difficile) or in other nosocomial pathogens, it is a fertile ground for future study.
Finally, efforts to decrease use of certain antibiotics might increase infection rates due to
inappropriate withholding of appropriate antibiotics. However, one reviewed study 10 noted no
increase in rates of surgical site infections following decrease in the use of vancomycin for
preoperative prophylaxis (see also Subchapter 20.1).
143

Costs and Implementation
The costs of implementing a program to alter antibiotic use practices must be balanced
against potential cost savings. Sources of savings may be reduced antibiotic use, use of less
expensive agents rather than the more expensive newer agents, and potentially, reduced costs
due to decreased incidence of nosocomial infections as a result of interventions. Although
several studies reported cost savings due only to decreased antibiotic use, 10,11,29 analyses taking
into account costs related to subsequent infections (or infections prevented) have been sparse.
One study noted that cost savings from decreased use of clindamycin offset the expenditures due
to increased use of other antibiotics. 7 The authors suggested that if each case of C. difficile
resulted in a cost of $2000, the savings to the hospital of the intervention could approach
$162,000 annually based on the number of cases averted. 7
Another cost of antibiotic use interventions is the expense of ongoing monitoring of
antibiotic use and antimicrobial susceptibilities of nosocomial pathogens. Effective
recommendation of certain antimicrobial agents over others requires access to (and financial and
logistic support for) routine antimicrobial susceptibility testing. Monitoring institutional
resistance patterns is vital in order to make required formulary changes in response to emerging
resistance patterns and to determine the most effective agents given prevailing susceptibility
patterns.
Comment
Given the strong association between antibiotic use and subsequent infection
(demonstrated for both C. difficile and VRE), it is not surprising that changes in antibiotic use
practices can reduce the incidence of infection with these 2 pathogens. The majority of reviewed
studies demonstrated a significant reduction in the incidence of VRE or C. difficile following
interventions to change antibiotic use practice. While these studies all demonstrated short-term
success, future studies should confirm the efficacy of such interventions over the long term. In
addition, the effectiveness and feasibility of combining antibiotic practice strategies with efforts
to enhance barrier precautions (Chapter 13) should be investigated. Finally, the costeffectiveness
of such strategies (taking into account both the costs associated with monitoring
and maintaining sound antibiotic use practices and the costs associated with nosocomial
antibiotic-resistant infections) should be investigated.
144

Table 14.1. Before-after studies of practices to improve antibiotic use*
Study Setting and Intervention Outcomes Results: before vs. after practice
Elderly care unit of a large teaching hospital in
England, 1984-85; Changes in empiric
antibiotic regimens 29
Chronic care facility in Baltimore, 1985-86;
multifaceted intervention 21
Veterans Affairs Medical Center in Arizona,
1990-92; restriction of clindamycin use 28
660-bed Veterans Affairs hospital in
California, 1992-94; removal of antibiotic
restrictions 30
703-bed Veterans Affairs Medical Center in
Virginia, 1993-94; restriction of clindamycin
use 7
557-bed academic medical center in Maryland,
1994; restriction of vancomycin use 8
35-bed hematologic malignancy unit in a large
medical center in England, 1994-95; sequential
antimicrobial formulary changes 9
Large academic medical center in Virginia,
1994-95; computer-based restriction of
vancomycin use 10
310-bed Veterans Affairs medical center in
New York, 1995; restriction of multiple
antibiotics 27
725-bed teaching hospital in Philadelphia,
1995-96; restriction of vancomycin use 11
Level 1 C. difficile infections decreased from 37
to 16 cases (p=0.002).
Level 1 Patients with C. difficile toxin decreased
from 28% to 24% (p=NS); Patients with
C. difficile culture increased from 33%
to 42% (p=NS)
Level 1 C. difficile infections decreased from
7.7 to 1.9 cases/month (p

References
1. Kunin CM, Tupasi T, Craig WA. Use of antibiotics. A brief exposition of the problem and
some tentative solutions. Ann Intern Med. 1973;79:555-560.
2. Gaynes R. The impact of antimicrobial use on the emergence of antimicrobial-resistant
bacteria in hospitals. Infect Dis Clin North Am. 1997;11:757-765.
3. Goldmann DA, Huskins WC. Control of nosocomial antimicrobial-resistant bacteria: a
strategic priority for hospitals worldwide. Clin Infect Dis. 1997;24:S139-145.
4. Murray BE. Vancomycin-resistant enterococcal infections. N Engl J Med. 2000;342:710-
721.
5. Johnson S, Gerding DN. Clostridium difficile-associated diarrhea. Clin Infect Dis.
1998;26:1027-1036.
6. Roghmann MC, McCarter RJ, Brewrink J, Cross AS, Morris JG. Clostridium difficile
infection is a risk factor for bacteremia due to vancomycin-resistant enterococci (VRE) in
VRE-colonized patients with acute leukemia. Clin Infect Dis. 1997;25:1056-1059.
7. Climo MW, Israel DS, Wong ES, Williams D, Coudron P, Markowitz SM. Hospital-wide
restriction of clindamycin: effect on the incidence of Clostridium difficile-associated
diarrhea and cost. Ann Intern Med. 1998;128:989-995.
8. Morris JG, Shay DK, Hebden JN, McCarter RJ, Perdue BE, Jarvis W, et al. Enterococci
resistant to multiple antimicrobial agents, including vancomycin. Establishment of
endemicity in a university medical center. Ann Intern Med. 1995;123:250-259.
9. Bradley SJ, Wilson ALT, Allen MC, Sher HA, Goldstone AH, Scott GM. The control of
hyperendemic glycopeptide-resistant Enterococcus spp. on a haematology unit by changing
antibiotic usage. J Antimicrob Chemother. 1999;43:261-266.
10. Anglim AM, Klym B, Byers KE, Scheld WM, Farr BM. Effect of a vancomycin restriction
policy on ordering practices during an outbreak of vancomycin-resistant Enterococcus
faecium. Arch Intern Med. 1997;157:1132-1136.
11. Morgan AS, Brennan PJ, Fishman NO. Impact of a vancomycin restriction policy on use
and cost of vancomycin and incidence of vancomycin-resistant enterococcus. Ann
Pharmacother. 1997;31:970-973.
12. Stosor V, Peterson LR, Postelnick M, Noskin GA. Enterococcus faecium bacteremia: does
vancomycin resistance make a difference? Arch Intern Med. 1998;158:522-527.
13. Clabots CR, Johnson S, Olson MM, Peterson LR, Gerding DN. Acquisition of Clostridium
difficile by hospitalized patients: evidence of colonized new admissions as the source of
infection. J Infect Dis. 1992;166:561-567.
14. Wilcox MH, Smyth ETM. Incidence and impact of Clostridium difficile infection in the
UK, 1993-1996. J Hosp Infect. 1998;39:181-187.
15. Gerding DN, Olson MM, Peterson LR, et al. Clostridium difficile-associated diarrhea and
colitis in adults: a prospective case-controlled epidemiologic study. Arch Intern Med.
1986;146:95-100.
16. Yablon SA, Krotenberg, Fruhmann K. Clostridium difficile-related disease: evaluation and
prevalence among inpatients with diarrhea in two freestanding rehabilitation hospitals.
Arch Phys Med Rehabil. 1993;74:913.
17. Olson MM, Shanholtzer CJ, Lee JT, Gerding DN. Ten years of prospective Clostridium
difficile-associated disease surveillance and treament at the Minneapolis VA Medical
Center, 1982-1991. Infect Control Hosp Epidemiol. 1994;15:371-381.
146

18. MacGowan AP, Brown I, Feeney R, Lovering A, McCulloch SY, Reeves DS. Clostridium
difficile associated diarrhea and legnth of hospital stay. J Hosp Infect. 1995;31:241-244.
19. Riley TV, Codde JP, Rouse IL. Increase length of stay due to Clostridium difficile
associated diarrhoea. Lancet. 1995;345:455-456.
20. Kent KC, Rubin MS, Wroblewski L, Hanff PA, Sline W. The impact of Clostridium
difficile on a surgical service. Ann Surg. 1998;227:296-301.
21. Bender BS, Laughon BE, Gaydos C, Forman MS, Bennett R, Greenough WB, et al. Is
Clostridium difficile endemic in chronic-care facilties? Lancet. 1986;2:11-13.
22. Spencer RC. Clinical impact and associated costs of Clostridium difficile-associated
disease. J Antimicrob Chemother. 1998;41(Suppl C):C5-C12.
23. Pestotnik SL, Classen DC, Evans RS. Implementing antibiotic practice guidelines through
computer-assisted decision support: clinical and financial outcomes. Ann Intern Med.
1996;124:884-890.
24. Yates RR. New intervention strategies for reducing antibiotic resistance. Chest.
1999;115:24S-27S.
25. Piquette RK. A targeted review of vancomycin use. Can J Hosp Pharm. 1991;44:83-7.
26. John JF, Fishman NO. Programmatic role of the infectious diseases physician in
controlling antimicrobial costs in the hospital. Clin Infect Dis. 1997;24:471-485.
27. Quale J, Landman D, Saurina G, Atwood E, DiTore V, Patel K. Manipulation of a hospital
antimicrobial formulary to control an outbreak of vancomycin-resistant enterococci. Clin
Infect Dis. 1996;23:1020-1025.
28. Pear SM, Williamson TH, Bettin KM, Gerding DN, Galgiani JN. Decrease in nosocomial
Clostridium difficile-associated diarrhea by restricting clindamycin use. Ann Intern Med.
1994;120:272-277.
29. McNulty C, Logan M, Donald IP, Ennis D, Taylor D, Baldwin RN, et al. Successful
control of Clostridium difficile infection in an elderly care unit through use of a restrictive
antibiotic policy. J Antimicrob Chemother. 1997;40:707-711.
30. Ho M, Yang D, Wyle FA, Mulligan ME. Increased incidence of Clostridium difficileassociated
diarrhea following decreased restriction of antibiotic use. Clin Infect Dis.
1996;23 (suppl 1):S102-S106.
31. Handwerger S, Raucher B, Altarac D, Monka J, Marchione S, Singh KV, et al. Nosocomial
outbreak due to Enterococcus faecium highly resistant to vancomycin, penicillin, and
gentamicin. Clin Infect Dis. 1993;16:750-755.
32. Chang VT, Nelson K. The role of physical proximity in nosocomial diarrhea. Clin Infect
Dis. 2000;31:717-722.
33. Mayfield JL, Leet T, Miller J, Mundy LM. Environmental control to reduce transmission of
Clostridium difficile. Clin Infect Dis. 2000;31:995-1000.
34. Byers KE, Durbin LJ, Simonton BM, Anglim AM, Adal KA, Farr BM. Disinfection of
hospital rooms contaminated with vancomycin-resistant Enterococcus faecium. Infect
Control Hosp Epidemiol. 1998;19:261-264.
35. Jacoby GA. Extended-spectrum beta-lactamases and other enzymes providing resistance to
oxyimino-beta-lactams. Infect Dis Clin North Am. 1997;11:875-887.
36. Michel M, Gutmann L. Methicillin-resistant Staphylococcus aureus and vancomycinresistant
enterococci: therapeutic realities and possibilities. Lancet. 1997;349:1901-1906.
37. Smith TL, Pearson ML, Wilcox KR, Cruz C, Lancaster MV, Robinson-Dunn B, et al.
Emergence of vancomycin resistance in Staphylococcus aureus. Glycopeptide-Intermediate
Staphylococcus aureus Working Group. N Engl J Med. 1999;340:493-501.
147

38. Archer GL. Staphylococcus aureus: a well-armed pathogen. Clin Infect Dis. 1998;26:1179-
1181.
39. Edmond MB, Wenzel RP, Pasculle AW. Vancomycin-resistant Staphylococcus aureus:
perspectives on measures needed for control. Ann Intern Med. 1996;124:329-334.
148

Chapter 15. Prevention of Nosocomial Urinary Tract Infections
Sanjay Saint, MD, MPH
University of Michigan School of Medicine
Background
Many hospitalized patients require the placement of indwelling urinary catheters for days
or even weeks at a time. 1 Only a minority of patients develop urinary tract infections because of
the presence of these devices, 2,3 but the frequency of their use produces substantial overall
morbidity for patients and costs to the health care system. Urinary tract infections (UTIs)
account for up to 40% of nosocomial infections, 4,5 with urinary catheter-related infections
causing the vast majority of nosocomial UTIs. 6 Each hospital-acquired UTI adds approximately
$675 to the costs of hospitalization. When bacteremia develops, this additional cost increases to
at least $2800. 2
Because of the substantial complications and costs associated with the use of urinary
catheters, a number of practices have been evaluated in an effort to reduce the incidence of
urinary catheter-related infections. This chapter reviews the evidence supporting the use of silver
alloy coated urinary catheters, and, because of its similarity, the recently described practice of
using urinary catheters impregnated with the antibiotic combination of minocycline and
rifampin. Subchapter 15.2 reviews the evidence supporting the use of suprapubic catheters as an
alternative to urethral catheters.
Subchapter 15.1. Use of Silver Alloy Urinary Catheters
Practice Description
Silver is a highly effective antibacterial substance, which can be applied to various types
of catheters. (See Subchapter 16.2 for a discussion of intravascular catheters coated with a
combination of silver sulfadiazine and chlorhexidine). Multiple studies have suggested that
silicone urethral catheters coated with hydrogel and silver salts reduce the risk of developing
bacteriuria, compared with standard latex urethral catheters (Foley catheters). As shown in a
recent meta-analysis, this benefit applies to catheters coated with silver alloy (which are coated
on both internal and external surfaces of the catheter), but not silver oxide (which are coated on
the external catheter surface only). Consequently, this chapter focuses only on studies evaluating
silver alloy catheters, and the use of catheters coated with antimicrobials. 8
Prevalence and Severity of the Target Safety Problem
Almost one million episodes of nosocomial UTI occur annually in the United States. 9
Each year approximately 96 million urethral catheters are sold worldwide. Of these, nearly 25%
are sold in the United States. 3 The daily rate of bacteriuria in catheterized patients ranges from 3
to 10%, with the incidence directly related to the duration of catheterization. 4 Among patients
with bacteriuria, 10 to 25% will develop symptoms of local urinary tract infection, 2,10 such as
suprapubic or flank pain. The development of catheter-related bacteriuria carries with it a 2.8fold
increased risk of death, independent of other co-morbid conditions and disease severity. 11,12
Bacteremia results from catheter-related bacteriuria in approximately 3% of patients, and
invariably represents a serious complication. 2,3
Beyond the morbidity and mortality associated with indwelling catheters, catheter-related
infection results in substantially increased health care costs. Data suggest that each episode of
149

hospital-acquired symptomatic catheter-related UTI costs an additional $676, and each episode
of catheter-related nosocomial bacteremia costs a minimum of $2836. 2
Estimates from one university hospital, based on data from almost 20 years ago, were
that hospital-acquired UTI led to approximately $204,000 in additional expenses per year. 13
More recent data are unavailable, but the institutional costs attributable to catheter-related
infection are clearly substantial.
Opportunities for Impact
Since catheter-related UTI is the leading cause of nosocomial infection in the United
States and is associated with increased morbidity and costs, any intervention that reduces the
incidence of catheter-related UTI is potentially important. Currently, it is unknown what
proportion of patients with indwelling catheters receives silver alloy catheters, however it is
likely to be the minority.
Study Designs
As shown in Table 15.1.1, a meta-analysis 7 which included 4 randomized clinical
trials, 14-17 compared the efficacy of silver catheters with standard, non-coated catheters. Five
additional studies 18-22 have appeared since publication of this meta-analysis. In 3 of these
studies, 18,20,22 the patient represented the unit of analysis. Another study employed a randomized
crossover design (Level 1), randomizing wards rather than individual patients. 19 The final study
used a prospective, before-after design at 5 different hospitals (Level 2). 21
The patient populations for these studies included patients on various hospital services
including urology, internal medicine, neurology, and the intensive care unit. In general, the
studies included patients expected to be catheterized for at least 2 days. Since the patients resided
in acute care hospitals rather than extended care centers, most were catheterized for 10 days or
less. Several studies specified that patients given concomitant antibiotics were excluded. 15-18
Study Outcomes
The individual trials and the meta-analysis focused primarily on the surrogate outcome of
bacteriuria (Level 2). The definition of bacteriuria varied somewhat in the studies. However,
low-level growth from a catheterized specimen (ie, 10 2 colony forming units (CFU) /mL) usually
progresses within days to concentrations of greater than 10 4 CFU/mL unless antibiotic therapy is
given. 23 Unfortunately, none of the studies was adequately powered to detect a significant
difference in the clinically more important outcomes of catheter-related bacteremia or death.
Though bacteriuria is a surrogate endpoint, 24 it is probably appropriate to use since it is a
component of the only causal pathway in the disease process between catheterization and an
important clinical outcome (eg, symptomatic UTI or catheter-related bacteremia). One study did
report differences in secondary bloodstream infections. 19
150

Evidence for Effectiveness of the Practice
The 4 clinical trials 14-17 of silver alloy catheters included in the meta-analysis 7 all showed
a significant reduction in the development of catheter-associated bacteriuria. As shown in Table
15.1.1, studies published after the meta-analysis have reported more mixed results. Several of the
studies have shown a statistically significant benefit of silver alloy catheters, but with a smaller
relative risk reduction compared to that reported in the meta-analysis. 19,21,22 However, one study
failed to find a significant benefit associated with silver alloy catheters, 20 and another found
benefit from silver alloy catheters in those given such catheters for about 5 days, but not in those
given the catheter for 14 days. 18 A formal update of the previous meta-analysis would be helpful,
but is beyond the scope of the current report.
Potential for Harm
There is likely minimal harm from the use of silver alloy urinary catheters. The one
theoretical harm involves the development of antimicrobial resistance. However, since silver is
not used systemically in the form of an antimicrobial agent for treatment, the clinical
significance of antimicrobial resistance to silver is unclear.
Costs and Implementation
Each silver alloy urinary catheter tray costs about $5.30 more than a standard, non-coated
urinary catheter tray. However, a recent economic evaluation indicates that when all the clinical
and economic costs are accounted for, silver alloy urinary catheters may provide both clinical
and economic benefits in patients receiving indwelling catheterization for 2 to 10 days. 3 It should
be noted that one of the major assumptions made in the economic evaluation is that a certain
proportion of patients with bacteriuria develop the clinically important (Level 1) outcomes of
symptomatic UTI or bacteremia. The economic analysis did not assign any costs to bacteriuria
but did assign costs if patients developed these clinically important outcomes. Additionally,
several of the very recent efficacy studies of silver alloy catheters 19,21,22 were not included in the
economic analysis. A clinical study, adequately powered to detect both meaningful clinical and
economic endpoints, would confirm the results of this economic evaluation that relied on
modeling techniques. The overall cost of universal implementation of silver alloy catheters is
unclear.
Comment
The data supporting the use of silver alloy urinary catheters to reduce urinary catheterrelated
bacteriuria is reasonably strong. As noted, the incidence of bacteriuria, while not
extremely high, carries a high morbidity. It remains unclear whether silver alloy urinary
catheters will also lead to decreases in the clinically more important outcomes of catheter-related
bacteremia and mortality. Continuing investigation into the impact of silver alloy catheters on
these important outcomes and their effect on the emergence of antibiotic resistance should be
pursued.
151

Of note, catheters coated with antibacterial substances other than silver have also been
evaluated. A recent randomized study 8 found that patients who received antimicrobialimpregnated
catheters coated with minocycline and rifampin had significantly lower rates of
gram-positive bacteriuria than a control group given standard, non-coated catheters (7.1% vs.
38.2%; p

Table 15.1.1. Studies of silver alloy and antibiotic-impregnated urethral catheters*
Study Description Design,
Outcomes
Saint,
1998 7
Maki,
1998 22
Verleyen,
1999 18
Bologna,
1999 21
Karchmer,
2000 19
Thibon,
2000 20
Meta-analysis of 4 randomized
controlled trials (n=453) of
silver alloy vs. uncoated urinary
catheters
Prospective, randomized,
double-blind trial of silver alloy
(n=407) vs. standard Foley
(n=443) catheters
Prospective, randomized study
of medium-term catheterization
with silver alloy (n=18) vs.
silicone (n=17) catheters after
radical prostatectomy
Prospective, randomized study
of short-term catheterization
with silver alloy (n=79) vs. latex
(n=101) catheters
Prospective, blinded study of
silver alloy vs. standard latex
Foley catheters in 5 hospitals.
Baseline period ranged from 3-
12 months (mean, 8 months);
intervention period ranged from
7-19 months (mean, 10 months)
12-month randomized crossover
trial of catheter-associated
urinary tract infections in
patients with silver-coated and
uncoated catheters. The ward
was the unit of analysis. A cost
analysis was also conducted.
Multicenter, prospective,
randomized, double-blind trial of
silver alloy (n=90) vs. standard
(n=109) catheters in patients
Level 1A,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 2,
Level 1
Level 1,
Level 1
Level 1,
Level 2
153
Results: Odds or Risk of Bacteriuria†
(unless otherwise noted)
OR 0.24 (95% CI: 0.11-0.52)
RR 0.74 (95% CI: 0.56-0.99)
After 14 days, 50.0% vs. 53.3% (p=NS)
On day 5, 6.3% vs. 11.9% (p

Darouiche,
1999 8
requiring catheterization for >3
days
Multicenter, prospective,
randomized, blinded trial of
medium-term catheterization
(mean, 14 days) with
minocycline-rifampin
impregnated (n=56) vs. silicone
(n=68) catheters after radical
prostatectomy
Level 1,
Level 2
154
Patients took longer to develop
bacteriuria with antimicrobialimpregnated
catheters than control
catheters (p=0.006 by the log-rank
test)
Overall bacteriuria at day 7: 15.2% vs.
39.7% (p

9. Kunin CM. Urinary Tract Infections: Detection, Prevention, and Management. 5th ed.
Baltimore: Williams and Wilkins; 1997.
10. Tambyah PA, Maki DG. Catheter-associated urinary tract infection is rarely symptomatic:
a prospective study of 1,497 catheterized patients. Arch Intern Med. 2000;160:678-682.
11. Platt R, Polk BF, Murdock B, Rosner B. Mortality associated with nosocomial urinary-tract
infection. N Engl J Med. 1982;307:637-642.
12. Platt R, Polk BF, Murdock B, Rosner B. Reduction of mortality associated with
nosocomial urinary tract infection. Lancet. 1983;1:893-897.
13. Krieger JN, Kaiser DL, Wenzel RP. Nosocomial urinary tract infections: secular trends,
treatment and economics in a university hospital. J Urol. 1983;130:102-106.
14. Lundeberg T. Prevention of catheter-associated urinary-tract infections by use of silverimpregnated
catheters [letter]. Lancet. 1986;2:1031.
15. Liedberg H, Lundeberg T. Silver alloy coated catheters reduce catheter-associated
bacteriuria. Br J Urol. 1990;65:379-381.
16. Liedberg H, Lundeberg T, Ekman P. Refinements in the coating of urethral catheters
reduces the incidence of catheter-associated bacteriuria. An experimental and clinical
study. Eur Urol. 1990;17:236-240.
17. Liedberg H, Lundeberg T. Prospective study of incidence of urinary tract infection in
patients catheterized with bard hydrogel and silver-coated catheters or bard hydrogelcoated
catheters [abstract]. J Urol. 1993;149:405A.
18. Verleyen P, De Ridder D, Van Poppel H, Baert L. Clinical application of the Bardex IC
Foley catheter. Eur Urol. 1999;36:240-246.
19. Karchmer TB, Giannetta ET, Muto CA, Strain BA, Farr BM. A randomized crossover
study of silver-coated urinary catheters in hospitalized patients. Arch Intern Med.
2000;160:3294-3298.
20. Thibon P, Le Coutour X, Leroyer R, Fabry J. Randomized multi-centre trial of the effects
of a catheter coated with hydrogel and silver salts on the incidence of hospital-acquired
urinary tract infections. J Hosp Infect. 2000;45:117-124.
21. Bologna RA, Tu LM, Polansky M, Fraimow HD, Gordon DA, Whitmore KE.
Hydrogel/silver ion-coated urinary catheter reduces nosocomial urinary tract infection rates
in intensive care unit patients: a multicenter study. Urology. 1999;54:982-987.
22. Maki DG, Knasinski V, Halvorson K, Tambyah PA. A novel silver-hydrogel-impregnated
indwelling urinary catheter reduces CAUTIs: a prospective double-blind trial [abstract].
Infect Control Hosp Epidemiol. 1998;19:682(A10).
23. Stark RP, Maki DG. Bacteriuria in the catheterized patient. What quantitative level of
bacteriuria is relevant? N Engl J Med. 1984;311:560-564.
24. Fleming TR, DeMets DL. Surrogate end points in clinical trials: are we being misled? Ann
Intern Med. 1996;125:605-613.
Subchapter 15.2. Use of Suprapubic Catheters
Background
As discussed in Subchapter 15.1, the use of indwelling urethral catheters results in
substantial morbidity and mortality. Given the medical and social morbidity associated with
urethral catheters, many clinicians have considered suprapubic catheterization as an alternative
155

to catheterization via the urethra. Suprapubic catheters are inserted in the lower abdomen, an
area with less bacterial colonization than the periurethral region, so that the risk for infection is
thought to be lower than with urethral catheters. Furthermore, although the suprapubic placement
of urinary catheters represents a minor surgical procedure, patients may find the result more
comfortable89 and, as reviewed below, the development of infectious complications is reduced.
Subchapter 15.1 discusses the use of silver alloy urinary catheters. The focus of this chapter is
the use of suprapubic catheters as compared with standard urethral indwelling catheters in adults.
Practice Description
Suprapubic catheterization typically involves the percutaneous placement of a standard
urinary catheter directly into the bladder. The procedure is performed by urologists using sterile
technique. It is generally performed in the operating room and is considered minor surgery.
Prevalence and Severity of Target Problem
In addition to the infectious complications (and their associated costs) discussed in
Subchapter 15.1, the use of urethral catheters causes substantial patient discomfort. In a recent
study at a Veteran Affairs Medical Center, 42% of catheterized patients surveyed reported that
the indwelling catheter was uncomfortable, 48% complained that it was painful, and 61% noted
that it restricted their activities of daily living. 7 Restricted activity reduces patient autonomy and
may promote other nosocomial complications, such as venous thromboembolism and pressure
ulcers. In addition, 30% of survey respondents stated that the catheter’s presence was
embarrassing, and in unsolicited comments that supplemented the structured questionnaires
several noted that it “hurts like hell.” 7
Opportunities for Impact
Since catheter-related urinary tract infection (UTI) is the leading cause of nosocomial
infection in the United States and is associated with increased morbidity and costs, any
intervention that reduces the incidence of catheter-related UTI is potentially important.
Currently, it is unknown what proportion of patients who require indwelling urinary catheters
receive suprapubic catheters, however, this practice is uncommon.
Study Design
There have been twelve prospective studies, 8,9,11-17 all but one randomized, 15 comparing
the efficacy of suprapubic catheters with standard, non-coated catheters (Table 15.2.1). In all of
these studies, the patient was the unit of analysis. The patient populations for these studies varied
but generally included patients with acute urinary retention and those undergoing various
surgical procedures. Since most of the patients evaluated resided in acute care hospitals, the
average duration of catheterization was generally less than 14 days.
Study Outcomes
All the trials focused on the outcome of bacteriuria. Several of the studies also assessed
patient satisfaction and the incidence of mechanical complications. The definition of bacteriuria
varied somewhat in the studies. However, low-level growth from a catheterized specimen (ie,
10 2 colony forming units (CFU)/mL) usually progresses within days to concentrations of greater
than 10 4 CFU/mL, unless antibiotic therapy is given. 18 Unfortunately, none of the studies was
adequately powered to detect a significant difference in the clinically more important outcomes
of catheter-related bacteremia or death. Though bacteriuria is a surrogate endpoint, 19 it is
156

probably appropriate to use since it is a component of the only causal pathway in the disease
process between suprapubic catheterization and an important clinical outcome (eg, symptomatic
UTI or catheter-related bacteremia).
Evidence for Effectiveness of the Practice
As shown in Table 15.2.1, studies comparing suprapubic catheterization with urethral
catheterization have produced mixed results. 8,9,11-17,20-22 Six trials reported lower rates of
bacteriuria in patients with suprapubic catheters, 11,13,15,16,21,22 and 4 trials indicated greater patient
satisfaction with suprapubic as opposed to urethral catheters. 8,13,16,20 In 3 of the studies, however,
mechanical complications were higher in those receiving suprapubic catheters. 12,15,16 Of note, 3
studies found that patients given suprapubic catheters have significantly decreased incidence of
urethral strictures compared with patients who received urethral catheters. 15,23,24 However, the
use of prophylactic antibiotics in patients receiving urethral catheters for transurethral resection
of the prostate has been shown to significantly reduce the incidence of strictures in the anterior
urethra. 25
Potential for Harm
As stated above, the primary problem associated with suprapubic catheter use involves
mechanical complications associated with insertion, most commonly catheter dislodgement or
obstruction, and failed introduction. The safe insertion of suprapubic indwelling urinary
catheters depends on trained personnel.
Costs and Implementation
The cost of each suprapubic urinary catheter tray is comparable to the cost of each
standard, non-coated urethral catheter tray. However, the overall initial costs of using suprapubic
catheters will no doubt be greater since procedure-related costs are substantially higher for
suprapubic than urethral catheters. Nurses are able to place urethral catheters at the bedside, but
urologists must place suprapubic catheters, and the procedure typically occurs in the operating
room. Additionally, it is unclear whether urologists are currently proficient at the insertion of
suprapubic catheters given how infrequently they are used. If suprapubic catheters are shown to
be effective, they may have a positive impact on patient care. The cost of training individuals in
inserting and maintaining the suprapubic catheter is likely to be substantial.
Comment
When compared with standard urethral indwelling catheters, suprapubic urinary catheters
may reduce urinary catheter-related bacteriuria. Additionally, patient satisfaction may be greater
with suprapubic catheters, although there is also evidence that patients placed with suprapubic
catheters more frequently experience certain mechanical complications. On the other hand,
urethral catheters are likely to lead to a higher incidence of urethral strictures. Given these mixed
results, conclusions regarding the overall benefit of routine suprapubic catheterization cannot
currently be made. However, it would be reasonable to consider conducting a formal metaanalysis
of the published trials to answer the question, “Compared with urethral indwelling
catheters, are suprapubic catheters less likely to lead to UTI (as measured by bacteriuria) and
more likely to lead to enhanced patient satisfaction?” Using explicit inclusion criteria and
accepted quantitative methods, a meta-analysis 26-28 can often help clarify the features of
individual studies that have divergent results. 29 In addition, a possible interaction between gender
of the patient and type of catheter is of interest since different pathophysiologic mechanisms
157

underlie the development of urethral catheter-related infection in men and women. 30 The
possibility of adequately evaluating effects within subgroups (eg, those undergoing certain
surgical procedures) because of an increased sample size is one of the benefits of metaanalysis.
31
If formal meta-analysis suggests that suprapubic catheters are less likely to lead to
urinary tract infection and more likely to enhance patient satisfaction, at least in some clinical
settings, then these catheters should be considered in the management of certain patients. On the
other hand, if the meta-analysis finds that urethral catheters are superior to suprapubic catheters,
then use of suprapubic catheters, albeit currently quite limited, should be further reduced.
158

Table 15.2.1. Prospective studies comparing suprapubic with urethral catheters
Study Design, Patient Bacteriuria (%)† Odds Ratio Comments§
Outcomes Population* Suprapubic Urethral (95% CI)‡
Shapiro,
1982 16
Andersen,
1985 13
Ichsan,
1987 9
Sethia,
1987 11
Schiotz,
1989 12
Horgan,
1992 15
O’Kelley,
1995 8
Ratnaval,
1996 14
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 2,
Level 2
Level 1,
Level 2
Level 1,
Level 2
General
surgical
patients with
urinary
retention
Women
undergoing
vaginal
surgery
Patients with
acute urinary
retention
General
surgical
patients
requiring
urine output
monitoring
Women
undergoing
vaginal
surgery
Men with
acute urinary
retention due
to prostatic
enlargement
General
surgical
patients
requiring
abdominal
surgery
Men
undergoing
colorectal
surgery
2/25 (8) 21/31
(68)
10/48 (21) 20/44
(45)
3/29 (10) 11/37
(30)
2/32 (6) 16/34
(47)
8/38 (21) 5/40
(12)
10/56 (18) 12/30
(40)
3/28 (11) 3/29
(10)
1/24 (4) 3/26
(12)
159
0.04
(0.01-0.24)
0.32
(0.11-0.86)
0.27
(0.04-1.22)
0.08
(0.01-0.41)
1.87
(0.48-8.01)
0.33
(0.11-0.99)
1.04
(0.13-8.51)
0.33
(0.01-4.60)
Pseudorandomized (urethral
catheters used in every third
patient) study; suprapubic
group had less pain but more
mechanical complications
Patients rated acceptability of
suprapubic catheters greater
None of the suprapubic group
complained of discomfort
compared with 17 of the
patients given urethral
catheters
Decrease in bacteriuria was
more significant in women
than in men
26% of suprapubic group
versus 5% of urethral group
had mechanical
complications
21% of suprapubic group
versus 3% of urethral group
had dislodgement; 0% of
suprapubic group versus 17%
of urethral group developed
urethral strictures
Study design unclear, but
probably not randomized;
suprapubic catheters caused
significantly fewer days of
catheter-related pain
Suprapubic group had fewer
voiding difficulties

Bergman,
1987 21
Abrams,
1980 20
Vandoni,
1994 22
Perrin,
1997 17
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Women
undergoing
vaginal
surgery for
stress
incontinence
Men with
urinary
retention
Patients
requiring
surgery for
various
indications
Patients
undergoing
rectal surgery
4/24 (17) 17/27
(63)
21/52 (40) 13/50
(26)
0/19 (0) 6/24
(25)
12/49 (24) 29/59
(49)
160
0.26
(0.10-0.68)
1.6
(0.88-2.75)
0
(0-0.95)
0.34
(0.13-0.83)
Length of hospital stay was
significantly less (by 1 day)
in the suprapubic catheter
group
12% of suprapubic catheter
group found catheter
uncomfortable compared with
64% in the standard urethral
catheter group (p

6. Saint S, Veenstra DL, Sullivan SD, Chenoweth C, Fendrick AM. The potential clinical and
economic benefits of silver alloy urinary catheters in preventing urinary tract infection.
Arch Intern Med 2000;160:2670-2675.
7. Saint S, Lipsky BA, Baker PD, McDonald LL, Ossenkop K. Urinary catheters: What type
do men and their nurses prefer? J Am Geriatr Soc 1999;47:1453-1457.
8. O'Kelly TJ, Mathew A, Ross S, Munro A. Optimum method for urinary drainage in major
abdominal surgery: a prospective randomized trial of suprapubic versus urethral
catheterization. Br J Surg 1995;82:1367-1368.
9. Ichsan J, Hunt DR. Suprapubic catheters: a comparison of suprapubic versus urethral
catheters in the treatment of acute urinary retention. Aust N Z J Surg 1987;57:33-36.
10. Krieger JN, Kaiser DL, Wenzel RP. Nosocomial urinary tract infections: secular trends,
treatment and economics in a university hospital. J Urol 1983;130:102-106.
11. Sethia KK, Selkon JB, Berry AR, Turner CM, Kettlewell MG, Gough MH. Prospective
randomized controlled trial of urethral versus suprapubic catheterization. Br J Surg
1987;74:624-625.
12. Schiotz HA, Malme PA, Tanbo TG. Urinary tract infections and asymptomatic bacteriuria
after vaginal plastic surgery. A comparison of suprapubic and transurethral catheters. Acta
Obstet Gynecol Scand 1989;68:453-455.
13. Andersen JT, Heisterberg L, Hebjorn S, Petersen K, Stampe Sorensen S, Fischer
Rasmussen W, et al. Suprapubic versus transurethral bladder drainage after
colposuspension/vaginal repair. Acta Obstet Gynecol Scand 1985;64:139-143.
14. Ratnaval CD, Renwick P, Farouk R, Monson JR, Lee PW. Suprapubic versus transurethral
catheterisation of males undergoing pelvic colorectal surgery. Int J Colorectal Dis
1996;11:177-179.
15. Horgan AF, Prasad B, Waldron DJ, O'Sullivan DC. Acute urinary retention. Comparison of
suprapubic and urethral catheterisation. Br J Urol 1992;70:149-151.
16. Shapiro J, Hoffmann J, Jersky J. A comparison of suprapubic and transurethral drainage for
postoperative urinary retention in general surgical patients. Acta Chir Scand 1982;148:323-
327.
17. Perrin LC, Penfold C, McLeish A. A prospective randomized controlled trial comparing
suprapubic with urethral catheterization in rectal surgery. Aust N Z J Surg 1997;67:554-
556.
18. Stark RP, Maki DG. Bacteriuria in the catheterized patient. What quantitative level of
bacteriuria is relevant? N Engl J Med 1984;311:560-564.
19. Fleming TR, DeMets DL. Surrogate end points in clinical trials: are we being misled? Ann
Intern Med 1996;125:605-613.
20. Abrams PH, Shah PJ, Gaches CG, Ashken MH, Green NA. Role of suprapubic
catheterization in retention of urine. J R Soc Med 1980;73:845-848.
21. Bergman A, Matthews L, Ballard CA, Roy S. Suprapubic versus transurethral bladder
drainage after surgery for stress urinary incontinence. Obstetrics & Gynecology
1987;69:546-549.
22. Vandoni RE, Lironi A, Tschantz P. Bacteriuria during urinary tract catheterization:
suprapubic versus urethral route: a prospective randomized trial. Acta Chir Belg
1994;94:12-16.
23. Hammarsten J, Lindqvist K, Sunzel H. Urethral strictures following transurethral resection
of the prostate. The role of the catheter. Br J Urol 1989;63:397-400.
161

24. Hammarsten J, Lindqvist K. Suprapubic catheter following transurethral resection of the
prostate: a way to decrease the number of urethral strictures and improve the outcome of
operations. J Urol 1992;147:648-651.
25. Hammarsten J, Lindqvist K. Norfloxacin as prophylaxis against urethral strictures
following transurethral resection of the prostate: an open, prospective, randomized study. J
Urol 1993;150:1722-1724.
26. Yusuf S. Obtaining medically meaningful answers from an overview of randomized
clinical trials. Stat Med 1987;6:281-294.
27. Light RJ. Accumulating evidence from independent studies: what we can win and what we
can lose. Stat Med 1987;6:221-231.
28. Hennekens CH, Buring JE, Hebert PR. Implications of overviews of randomized trials. Stat
Med 1987;6:397-409.
29. Sacks HS, Berrier J, Reitman D, Ancona Berk VA, Chalmers TC. Meta-analyses of
randomized controlled trials. N Engl J Med 1987;316:450-455.
30. Daifuku R, Stamm WE. Association of rectal and urethral colonization with urinary tract
infection in patients with indwelling catheters. JAMA 1984;252:2028-2030.
31. Gelber RD, Goldhirsch A. Interpretation of results from subset analyses within overviews
of randomized clinical trials. Stat Med 1987;6:371-388.
162

Chapter 16. Prevention of Intravascular Catheter-Associated Infections
Sanjay Saint, MD, MPH
University of Michigan School of Medicine
Background
Central venous catheters inserted for short-term use have become common and important
devices in caring for hospitalized patients, especially the critically ill. 1 While they have
important advantages (eg, ability to administer large volumes of fluid), short-term vascular
catheters are also associated with serious complications, the most common of which is infection.
Intravascular catheters are one of the most common causes of nosocomial bacteremia; 2 and
catheter-related bloodstream infection (CR-BSI) affects over 200,000 patients per year in the
United States. 3 This chapter focuses primarily on short-term central venous catheters. Two
relatively recent reviews address prevention of infection due to other types of vascular
catheters. 4,5 We review use of maximum barrier precautions (Subchapter 16.1), central venous
catheters coated with antibacterial or antiseptic agents (Subchapter 16.2), and use of
chlorhexidine gluconate at the insertion site (Subchapter 16.3). We review several promising
practices, as well as some common ineffective practices (Subchapter 16.4).
Definitions and Microbiology
Catheter-related infections can be subdivided into those that are local and those that are
bacteremic. Local infection involves only the insertion site and manifests as pericatheter skin
inflammation. Local infection is usually diagnosed when there is evidence of an insertion-site
infection (eg, purulence at the exit site). Catheter colonization is defined by growth of an
organism from the tip or the subcutaneous segment of the removed catheter. Growth of greater
than 15 colony-forming units (CFU) using the semiquantitative roll-plate culture technique is
often used to define catheter colonization. 6 Alternatively, the presence of more than 1000 CFUs
per catheter tip segment by quantitative culture using a method such as sonication indicates
evidence of catheter colonization. 7 Signs of local infection may or may not be present when there
is significant catheter colonization; evidence of local infection is observed in at least 5% of
patients with catheter colonization.
Bacteremic catheter-related infection (often also referred to as CR-BSI) is defined as a
positive blood culture with clinical or microbiologic evidence that strongly implicates the
catheter as the source of infection. 1 This includes: 1) evidence of local infection with isolation of
the same organism from both pus around the site and bloodstream; or 2) positive cultures of both
the catheter tip (using either semi-quantitative or quantitative methods) and bloodstream with the
same organism; or 3) clinical evidence of sepsis (eg, fever, altered mental status, hypotension,
leukocytosis) that does not respond to antibiotic therapy, but resolves once the catheter is
removed. 1,5 Some have proposed additional methods of diagnosing CR-BSI, including paired
blood cultures (drawn from both the central venous catheter and from a noncatheterized vein) 8
and a technique in which time to culture positivity for blood drawn from the central venous
catheter is compared with that for the blood drawn from percutaneous venipuncture. 9
The most common organisms causing catheter-related infections are staphylococci, gram
negative rods, and Candida species. 10,11 The pathophysiology of these infections include several
mechanisms, the most important of which involve the skin insertion site and the catheter hub. 1
Bacteria migrate from the insertion site on the skin along the external surface of the catheter and
163

then colonize the distal tip. 12,13 The hub can also lead to infection when bacteria are introduced
via the hands of medical personnel. These organisms then migrate along the internal surface of
the lumen and may result in bacteremia. 14
Less commonly, catheter-related infection can result from hematogenous seeding of the
catheter from another focus 15 or from contaminated infusates. 16
Prevalence and Severity of the Target Safety Problem
A recent quantitative review found that of patients in whom standard, non-coated central
venous catheters are in place on average for 8 days, 25% can be expected to develop catheter
colonization and 5% will develop CR-BSI. 17 The risk of CR-BSI from this estimate is similar to
the rate reported by the Federal Centers for Disease Control and Prevention (CDC). The CDC
has reported an average CR-BSI rate of 2.8 to 12.8 infections per 1000 catheter-days for all types
of intensive care units and average rates of 4.5 to 6.1 infections per 1000 catheter-days for
medical/surgical intensive care units. 18
CR-BSI is associated with an increased risk of dying, but whether this association is
causal remains controversial. 17 Some argue that hospitalized patients who develop CR-BSI may
differ in their clinical and physiologic characteristics, and thus may have a higher risk of dying
due to intrinsic factors. Proponents of this view believe that the development of CR-BSI is
primarily a marker of severe underlying disease or deficient immunity rather than an
independent risk factor for dying. Unfortunately, the few studies evaluating attributable mortality
due to CR-BSI have conflicting results.
Pittet and colleagues estimated that the attributable mortality of CR-BSI was 25% in a
matched case-control study. 19,20 Another matched study estimated that the attributable mortality
was 28%. 21 Other investigators have found a much smaller attributable mortality associated with
CR-BSI. DiGiovine et al, in a matched case-control study of 136 medical intensive care unit
patients, found a non-significant attributable mortality of CR-BSI (4.4%; p=0.51). 22 A recent,
carefully matched cohort study of 113 patients by Soufir and colleagues also failed to detect a
statistically significant increase in mortality associated with CR-BSI. 23 Nevertheless, given the
small sample size, these authors concluded that their findings are consistent with a 10% to 20%
increased mortality due to CR-BSI. 23 Further research to clarify the mortality associated with
CR-BSI is needed, but the available data are consistent with an attributable mortality of CR-BSI
ranging between 4% and 20%.
Central venous catheter related infection also leads to increased health care costs. Though
there is substantial variability in the economic estimates, a recent review estimates that an
episode of local catheter-related infection leads to an additional cost of approximately $400,
while the additional cost of CR-BSI ranges from about $6005 to $9738. 17 Some have estimated
that each episode leads to even higher costs, approximately $25,000 per episode. 19,20
164

Prevention
Unnecessarily prolonged catheterization should be avoided. Because of the increased risk
of infection with prolonged catheterization, many clinicians attempt to reduce this risk with
routine changes of the catheter, either over a guidewire or with a new insertion site. However,
the available data do not support this practice. 24 Eyer et al 25 randomized 112 surgical patients
receiving a central venous, pulmonary arterial, or systemic arterial catheter for more than 7 days
into three groups: a) weekly catheter change at a new site; or b) weekly guidewire exchange at
the same site; or c) no routine weekly changes. No significant difference was noted in the
incidence of local or bacteremic infection. 25 Cobb and colleagues 26 randomized 160 patients with
central venous or pulmonary arterial catheters to either replacement every 3 days at a new site or
over a guidewire, or replacement only when clinically indicated. In those with replacement
catheters at new sites, the risk of infectious complications was not decreased and the number of
mechanical complications was increased. Those undergoing routine replacement via a guidewire
exchange showed a trend towards a higher rate of bloodstream infections compared with those
who had catheter replacement only when clinically indicated. 26 A recent meta-analysis has
confirmed that routine changes of central venous and systemic arterial catheters appear
unnecessary; 24 attempts should be made, however, to limit the duration of catheterization. Strict
adherence to proper handwashing and use of proven infection control principles is crucial (see
Chapters 13 and 14). 27,28
Subchapter 16.1. Use of Maximum Barrier Precautions During Central Venous Catheter
Insertion
Practice Description
Catheter-related infections often result from contamination of the central venous catheter
during insertion. Maximum sterile barrier (MSB) precautions may reduce the incidence of
catheter contamination during insertion and thus reduce the rate of CR-BSI. MSB precautions
consist of the use of sterile gloves, long-sleeved gowns, and a full-size drape as well as a nonsterile
mask (and often a non-sterile cap) during central venous catheter insertion.
Opportunities for Impact
The proportion of patients receiving central venous catheters in whom maximum barrier
precautions are employed is not currently known. If maximum barrier precautions are not used,
then the standard insertion technique involves the use of only sterile gloves and a sterile small
drape. Given the additional time required to employ MSB, it is likely that many patients are not
receiving maximum barrier precautions during catheter insertion.
Study Designs
One randomized and one non-randomized study have evaluated the use of maximum
barrier precautions (Table 16.1.1). The clinical trial randomized 176 patients to catheter insertion
using MSB and 167 patients to control (use of sterile gloves and sterile small drape). 29 A nonrandomized
before-after observational evaluation assessed the effect of a 1-day course on
infection control practices and procedures on physician compliance with MSB use and incidence
of catheter-infection. 30
165

Study Outcomes
Both studies evaluated rates of catheter-related infection (Level 1), including local and
bloodstream infection.
Evidence for Effectiveness of the Practice
There is moderately strong evidence that use of maximum barrier precautions decrease
the risk of catheter-related infection (Table 16.1.1). Furthermore, the evidence that health care
providers—specifically physicians-in-training—can be taught proper use of barrier precautions
and thereby decrease the incidence of infection is reasonably strong.
Potential for Harm
There is virtually no harm associated with this intervention.
Costs and Implementation
The use of maximum barrier precautions will cost more than not using this technique in
both materials and time. Additionally, teaching health care providers how to properly use
maximum barrier precautions is also time-consuming and expensive. Sherertz and colleagues
estimated the overall cost of their educational program and supplies to be $74,081. 30 However,
when the costs of preventing catheter-related infection are also included, use of MSB has been
estimated to be cost-saving in simplified “back-of-the-envelope” cost studies. 29,30 Formal
economic evaluation is required to fully assess the economic consequences of full adoption of
maximum barrier precautions.
Comments
Use of MSB appears to be a reasonable method of preventing catheter-related infection.
Though achieving full compliance with this method of catheter insertion is likely to be
challenging, a relatively simple educational intervention has demonstrated effectiveness in
improving adherence and reducing infection rates. Given the excellent benefit-to-harm ratio of
this patient safety practice, it seems reasonable to strongly consider employing MSB for all
patients requiring central venous catheters. The economic consequences of full implementation
of this practice are still not entirely clear.
166

Table 16.1.1. Studies of vascular catheter-related infection*
Study Description; Intervention Study
Design,
Outcomes
343 patients in a 500-bed cancer referral
center; catheters inserted under maximal
sterile barrier precautions (mask, cap,
sterile gloves, gown, and large drape)
vs. control precautions (sterile gloves
and small drape only) 29
6 ICUs and a step-down unit in an
academic medical center in NC; 1-day
course for physicians-in-training on the
control of vascular catheter infection,
emphasizing use of full-size sterile
drapes 30
Meta-analysis of 12 RCTs (2611
catheters) comparing central venous
catheters coated with
chlorhexidine/silver sulfadiazine with
standard, non-coated catheters 44
High-risk adult patients at 12 universityaffiliated
hospitals in whom central
venous catheters were expected to
remain in place for 3 days;
minocycline/rifampin vs.
chlorhexidine/silver sulfadiazine
catheters 46
Meta-analysis of 7 RCTs (772 catheters)
comparing tunneling with standard
placement of short-term central venous
catheters 61
Meta-analysis of 12 RCTs comparing
prophylactic heparin use (in different
forms) with no heparin use on the
following outcomes: central venous
catheter colonization (3 trials), CR-BSI
(4 trials), and catheter-related deep
venous thrombosis (7 trials) 59
Level 1,
Level 1
Level 2‡,
Level 1
Level 1A,
Level 1
Level 1,
Level 1
Level 1A,
Level 1
Level 1A,
Level 1
167
Results (p-value or 95% CI)†
CR-BSI per 1000 catheter days: 0.08
vs. 0.5, (p=0.02)
Catheter colonization: 2.3% vs. 7.2%
(p=0.04)
Primary bloodstream infection and
catheter-related infection decreased
28% (p

Meta-analysis of 12 RCTs (918 patients,
1913 catheters) assessing the effect of
guidewire exchange and a prophylactic
replacement strategy (change every 3
days) on central venous catheter-related
colonization (8 trials), exit site infection
(4 trials), bacteremia (8 trials), and
mechanical complications (9 trials) in
critically ill patients 24
Level 1A,
Level 1
168
Catheter colonization: RR 1.26 (0.87-
1.84)
Exit site infection: RR 1.52 (0.34-6.73)
Bacteremia: RR 1.72 (0.89-3.33)
Mechanical complications: RR 0.48
(0.12-1.91)
Prophylactic catheter replacement
every 3 days was not found to be
better than as-needed replacement
* CI indicates confidence interval; CR-BSI, catheter-related bloodstream infection; OR, odds
ratio; RCT, randomized controlled trial; and RR, relative risk.
† Results are reported as intervention group vs. control (standard or usual care) group.
‡ Prospective before-after study design.

Subchapter 16.2. Use of Central Venous Catheters Coated with Antibacterial or
Antiseptic Agents
Practice Description
Recent studies have indicated that central venous catheters coated with antimicrobial
agents reduce the incidence of catheter-related bloodstream infection (CR-BSI). Implementing
use of these catheters would be simple, primarily involving the replacement of standard, noncoated
vascular catheters. However, these catheters, such as chlorhexidine/silver sulfadiazineimpregnated
catheters and minocycline/rifampin-coated catheters, are more expensive than
standard catheters. Thus, the cost-effectiveness of these catheters needs to be considered by
decision makers.
Opportunities for Impact
Currently, it is not known precisely what proportion of patient who require central
venous catheterization receive an antimicrobial catheter, however, it is probably the minority of
patients.
Study Designs
Multiple randomized trials have compared chlorhexidine/silver sulfadiazine central
venous catheters with standard, non-coated central venous catheters. 31-43 In addition, a recent
meta-analysis used a fixed effects model to combine the results of these chlorhexidine/silver
sulfadiazine trials. 44 A large, multicenter study has compared minocycline/rifampin coated
catheters with non-coated, standard catheters. 45 Additionally, a recent multicenter randomized
trial of minocycline/rifampin versus chlorhexidine/silver sulfadiazine catheters has also been
reported. 46 The majority of the patients enrolled in the individual studies cited above had a
central venous catheter in place for 8 days on average (range of average duration, 5 to 11 days).
Details of the characteristics and results of the trials comparing central venous catheters coated
with chlorhexidine/silver sulfadiazine to control catheters are in Tables 16.2.1 and 16.2.2.
Study Outcomes
Most studies reported the incidence of catheter colonization and CR-BSI. Though the
precise outcome definitions in some of the studies varied, in general the definition of catheter
colonization and CR-BSI used in most of these studies was explicit and appropriate.
Evidence for Effectiveness of the Practice
The evidence for the efficacy of chlorhexidine/silver sulfadiazine catheters is fairly
substantial. The recent meta-analysis found a statistically significant decrease in the incidence of
CR-BSI (odds ratio 0.56, 95% CI: 0.37-0.84). 44 There is also reasonable evidence that
minocycline-rifampin catheters reduce the risk of CR-BSI compared with standard, non-coated
catheters. The recent randomized trial of minocycline/rifampin versus chlorhexidine/silver
sulfadiazine catheters found a significant and clinically important decrease in the incidence of
CR-BSI in the group of patients using minocycline/rifampin compared with chlorhexidine/silver
sulfadiazine catheters (0.3% vs. 3.4%, p

concern. Although there have been no reports of hypersensitivity reactions to
chlorhexidine/silver sulfadiazine impregnated central venous catheters in the United States (out
of more than 2.5 million sold), 13 cases of immediate hypersensitivity reactions have been
reported in Japan, including one potentially associated death. 47 There were 117,000 antisepticimpregnated
catheters sold in Japan before their use was halted because of these cases. 47 It is not
clear why there have been no reports of hypersensitivity reactions in the U.S; this heterogeneity
may be caused by a higher previous exposure of patients in Japan to chlorhexidine or by a
genetic predisposition.
Minocycline and rifampin are both occasionally used as systemic antimicrobial agents;
thus, their use on catheters raises the important theoretical issue of increased antimicrobial
resistance. At this time, there has been no conclusive evidence that antimicrobial resistance has
or will increase due to the use of these catheters.
Costs and Implementation
Formal and informal economic analyses indicate that central venous catheters coated with
antibacterial agents (such as chlorhexidine/silver sulfadiazine or minocycline/rifampin) are likely
to lead to both clinical and economic advantages in selected patients. In terms of formal
economic comparisons, a recent analysis compared chlorhexidine/silver sulfadiazine catheters to
standard catheters and found that chlorhexidine/silver sulfadiazine catheters lead to both clinical
and economic advantages in patients receiving central venous catheterization for 2 to 10 days
and who were considered high risk for infection (ie, critically ill or immunocompromised
patients). Specifically, the chlorhexidine/silver sulfadiazine catheters led to a significant decrease
in the incidence of CR-BSI and death, and a cost savings of approximately $200 per catheter
used. 47 Importantly, the risk of hypersensitivity reaction to the chlorhexidine/silver sulfadiazine
catheters was considered in the analysis, but had little effect on the overall clinical and economic
outcomes. 47
However, given the recently demonstrated efficacy of the minocycline/rifampin catheter
compared with the chlorhexidine/silver sulfadiazine catheter, 46 a formal cost-effectiveness
analysis comparing these two types of coated catheters is necessary. This is especially important
since the minocycline/rifampin catheter costs about $9 more per catheter than the
chlorhexidine/silver sulfadiazine catheter.
Implementation of either of these catheters would be straightforward. Stocking the
appropriate antimicrobial catheter in areas of the hospital that are likely to require such catheters
(eg, intensive care unit, operative room, hematology-oncology floor) would be a relatively
simple way of translating the research findings into actual practice.
170

Comment
In light of the substantial clinical and economic burden of catheter-related infection,
hospital personnel should adopt proven cost-effective methods to reduce this common and
important nosocomial complication. The bulk of the evidence supports the use of either
chlorhexidine/silver sulfadiazine or minocycline/rifampin central venous catheters rather than
standard (non-coated) catheters in high-risk patients requiring short-term central venous
catheterization (eg, for 2 to 10 days). Choosing between the 2 antimicrobial catheters requires a
formal cost-effectiveness analysis since the minocycline/rifampin catheter costs significantly
more than the chlorhexidine/silver sulfadiazine catheter. There are 2 primary issues that should
be addressed when comparing these catheters: the expected duration of catheterization and the
risk of antibiotic resistance to the patient, the hospital, and society. Though each
minocycline/rifampin catheter costs more than the chlorhexidine/silver sulfadiazine catheter,
using minocycline/rifampin catheters may actually result in cost-savings for at least some patient
populations given their improved overall efficacy. Of note, the improved efficacy of the
minocycline/rifampin catheters may be a result of coating both the internal and external surfaces
with these substances; the chlorhexidine/silver sulfadiazine catheters evaluated to date have only
had the external surface coated with the antiseptic combination.
171

Table 16.2.1. Characteristics of trials comparing central venous catheters coated with
chlorhexidine/silver sulfadiazine to control catheters*
Study Description Number of
Catheters
(Treatment,
Control)
Tennenberg 31 : 282 hospital patients
(137 treatment, 145 control) in
variety of settings; double- and
triple-lumen catheters without
exchanges over guidewires
Maki 32 : 158 ICU patients (72
treatment, 86 control); triple-lumen
catheters with catheter exchanges
over guidewires
van Heerden 33 §: 54 ICU patients (28
treatment, 26 control); triple-lumen
catheters without catheter exchanges
over guidewires
Hannan 34 : ICU patients; triple-lumen
catheters
Bach 35 §: 26 ICU patients (14
treatment, 12 control); triple-lumen
catheters without catheter exchanges
over guidewires
Bach 36 §: 133 surgical patients (116
treatment, 117 control); double- and
triple-lumen cathetes without
exchanges over guidewires
Heard 37 §: 111 SICU patients (107
treatment, 104 control); triple-lumen
catheters with exchanges over
guidewires
Collin 38 : 119 ER/ICU patients (58
treatment, 61 control); single-,
double-, and triple-lumen catheters
with exchanges over guidewires
Ciresi 39 §: 191 patients receiving
TPN (92 treatment, 99 control);
triple-lumen catheters with
exchanges over guidewires
Pemberton 40 : 72 patients receiving
TPN (32 treatment, 40 control);
triple-lumen catheters without
exchanges over guidewires
137,
145
208,
195
28,
26
68,
60
14,
12
116,
117
151,
157
98,
139
124,
127
32,
40
172
Mean Catheter
Duration in
Days
(Treatment,
Control)
5.1,
53
6.0,
6.0
6.6,
6.8
7,
8
7.0,
7.0
7.7,
7.7
8.5,
9
9.0,
7.3
9.6,
9.1
10,
11
Catheter
Colonization†
SQ (IV, SC,
>15 CFU)
SQ (IV, >15
CFU)
SQ (IV, >15
CFU)
SQ (IV, >10 3
CFU)
QN (IV, >10 3
CFU)
QN (IV, >10 3
CFU)
SQ (IV, SC,
>14 CFU)
SQ (IV, SC,
>15 CFU)
SQ (IV, SC,
>15 CFU)
Catheter-
Related
Bloodstream
Infection†
SO (IV, SC,
site), CS, NS
SO (>15 CFU,
IV, hub, inf)‡
NR
SO (IV, >10 3
CFU), NS
NR
SO (IV)
SO (IV, SC,
>4 CFU)
SO (IV, SC)
SO (IV, SC)
NR SO (IV), Res,
NS

Ramsay 41 §: 397 hospital patients
(199 treatment, 189 control) in a
variety of settings; triple-lumen
catheters without exchanges over
guidewires
Trazzera 42 §: 181 ICU/BMT patients
(99 treatment, 82 control); triplelumen
catheters with exchanges over
guidewires
George 43 : Transplant patients; triplelumen
catheters without exchanges
over guidewires
199,
189
123,
99
44,
35
173
10.9,
10.9
11.2,
6.7
SQ (IV, SC,
>15 CFU)
SQ (IV, >15
CFU)
NR SQ (IV, >5
CFU)
SO (IV, SC)
SO (IV, >15
CFU)
SO (IV)
* BMT indicates bone marrow transplant; CFU, colony forming units; CS, clinical signs of
systemic infection; ER, emergency room; ICU, intensive care unit; IV, intravascular catheter
segment; inf, catheter infusate; NR, not reported; NS, no other sources of infection; QN,
quantitative culture; Res, resolution of symptoms upon catheter removal; SC, subcutaneous
catheter segment; SICU, surgical intensive care unit; site, catheter insertion site; SO, same
organism isolated from blood and catheter; SQ, semi-quantitative culture; and TPN, total
parenteral nutrition.
† Catheter segments (or site) cultured and criteria for a positive culture are given in parenthesis
‡ Organism identity confirmed by restriction-fragment subtyping
§ Additional information provided by author (personal communications, 1/98-3/98)
Culture method reported as semiquantitative; criteria for culture growth suggests quantitative
method

Study
Table 16.2.2. Results of trials comparing central venous catheters coated with
chlorhexidine/silver sulfadiazine to control catheters*
Tennenberg 31
Maki 32
Catheter Colonization Catheter-related Bloodstream Infection
No. (%) Positive Odds Ratio
(95% CI)
No. (%) Positive
Treatment Control
Treatment Control
174
Odds Ratio
(95% CI)
8 (5.8%) 32 (22.1%) 0.22 (0.10-0.49) 5 (3.6%) 9 (6.2%) 0.57 (0.19-1.75)
28 (13.5%) 47 (24.1%) 0.49 (0.29-0.82) 2 (1.0%) 9 (4.6%) 0.20 (0.04-0.94)
van Heerden 33 † 4 (14.3%) 10 (38.5%) 0.27 (0.07-1.00) – – –
Hannan 34
22 (32.4%) 22 (36.7%) 0.83 (0.40-1.72) 5 (7.4%) 7 (11.7%) 0.60 (0.18-2.00)
Bach 35 † 0 (0%) 4 (33.3%) 0 (0-0.65) – – –
Bach 36 † 2 (1.7%) 16 (13.7%) 0.11 (0.02-0.49) 0 (0%) 3 (2.6%) 0 (0-1.28)
Heard 37 † 60 (39.7%) 82 (52.2%) 0.60 (0.38-0.95) 5 (3.3%) 6 (3.8%) 0.86 (0.26-2.89)
Collin 38
2 (2.0%) 25 (18.0%) 0.10 (0.02-0.41) 1 (1.0%) 4 (2.9%) 0.35 (0.04-3.16)
Ciresi 39 † 15 (12.1%) 21(16.5%) 0.69 (0.34-1.42) 13 (10.5%) 14 (11.0%) 0.95 (0.43-2.10)
Pemberton 40
– – – 2 (6.3%) 3 (7.5%) 0.82 (0.13-5.24)
Ramsay 41 † 45 (22.6%) 63 (33.3%) 0.58 (0.37-0.92) 1 (0.5%) 4 (2.1%) 0.23 (0.03-2.11)
Trazzera 42 † 16 (13.0%) 24 (24.2%) 0.47 (0.23-0.94) 4 (3.3%) 5 (5.1%) 0.63 (0.17-2.42)
George 43
10 (22.7%) 25 (71.4%) 0.12 (0.04-0.33) 1 (2.3%) 3 (8.6%) 0.25 (0.02-2.50)
* CI indicates confidence interval.
† Additional information provided by author (personal communications, 1/98-3/98)

Subchapter 16.3. Use of Chlorhexidine Gluconate at the Central Venous Catheter
Insertion Site
Practice Description
Microbial populations on the skin are routinely suppressed with antiseptic agents prior to
catheter insertion. Using an antiseptic solution for skin disinfection at the catheter insertion site
helps to prevent catheter-related infections. The physician uses an agent that has antimicrobial
properties to thoroughly cleanse the skin just prior to insertion of a central venous catheter. In
the United States, povidone-iodine (PI) is overwhelmingly the most commonly used agent for
this purpose. Recently, several studies have compared the efficacy of PI and chlorhexidine
gluconate (CHG) solutions in reducing vascular catheter-related infections.
Opportunities for Impact
If PI is the most commonly used agent for site disinfection in the United States even
though CHG may be superior, substantial opportunity exists for impact by switching to CHG.
Study Designs
The study characteristics of 6 randomized trials 48-53 comparing any type of CHG solution
with PI solution for vascular catheter site care are shown in Table 16.3.1. The mean duration of
catheterization for the CHG and PI groups was comparable in most of the studies. There was no
significant difference in the sites at which catheters were inserted between the CHG and PI
groups. Several formulations of CHG were used, including 9,12-14 an alcoholic solution and an
aqueous solution. All studies used 10% PI solution for the control arm.
Study Outcomes
All studies 48-53 evaluated catheter colonization (Level 2 outcome) and all but one 52
evaluated CR-BSI (Level 1 outcome). All studies evaluating CR-BSI as an outcome required the
recovery of the same microbial species from both the catheter segment and a blood culture.
Evidence for Effectiveness of the Practice
Most clinical trials have revealed that the use of CHG solution results in a significant
decrease in catheter colonization, but the evidence is not clear for CR-BSI (Table 16.3.2). Most
of the individual trials showed a trend in reducing CR-BSI incidence in patients using CHG
solution. The lack of significant results may be a result of insufficient statistical power in the
individual studies. A formal meta-analysis of the published trials would be valuable in assessing
the comparative efficacy of PI versus CHG for central venous catheter site disinfection. Using
explicit inclusion criteria and accepted quantitative methods, a meta-analysis 54-56 can often help
clarify the features of individual studies that have divergent results 57 and increase statistical
power since several small studies can be pooled. 58
Potential for Harm
Only one study reported adverse effects from the use of either antiseptic solution. Maki et
al 48 found erythema at the insertion site in 28.3% of catheters in the PI group and in 45.3% of
catheters in the CHG group (p=0.0002). However, there was no statistically significant
difference in erythema among these 2 groups and those patients whose site was disinfected with
alcohol. Hypersensitivity reactions to chlorhexidine-silver sulfadiazine impregnated central
175

venous catheters and to use of CHG for bathing have been reported. Hypersenstivity reactions
were not reported in any of the studies, but clinicians should be aware of such potential side
effects. Another concern is the development of bacterial resistance. However, there have been
few reports of bacterial resistance to CHG despite its widespread use for several decades.
Costs and Implementation
The cost of CHG is approximately twice that of PI with an absolute difference of $0.51
(approximately $0.92 versus $0.41 for a quantity sufficient to prepare a central venous catheter
insertion site). If meta-analysis suggests that CHG use is effective in reducing the risk of CR-
BSI, a formal economic evaluation of this issue is required.
Comment
The use of chlorhexidine gluconate rather than povidone-iodine solution for catheter site
care may be an effective and simple measure for improving patient safety by reducing vascular
catheter-related infections. Formal meta-analysis and economic evaluations are required before
strongly recommending that CHG replace PI for central venous catheter site disinfection in
appropriate patient populations.
176

Table 16.3.1. Characteristics of studies comparing chlorhexidine gluconate (CHG) and
povidone-iodine (PI) solutions for vascular catheter site care*
Study Description† Number of
Catheters
(Treatment,
Control)
Maki 48 : 441 ICU
patients (2% aqueous
CHG solution in 214, PI
in 227)
Sheehan 49 : 189 ICU
patients (2% aqueous
CHG solution in 94, PI
in 95)
Meffre 50 : 1117 hospital
patients (CHG solution
of 0.5% alcohol 70% in
568, PI in 549)
Mimoz 51 : ICU patients
(Biseptine®§ vs. PI)
Cobett and LeBlanc 52 :
244 hospital patients
(0.5% alcohol 70% in 8,
PI in 161)
Humar et al 53 : 3374 ICU
patients (0.5% alcohol
in 193 and 181/193
Mean
Catheter
Duration in
Days
(Treatment,
Control)
177
Catheter
Colonization‡
Catheter-
Related
Bloodstream
Infection‡
214, 227 5.3, 5.3 SQ (>15 CFU) CX, NoSource,
Sx
169,177 NA SQ (>15 CFU) CX, NoSource,
Sx
568, 549 1.6, 1.6 SQ (>15 CFU) or
QN (>10 3
CFU/mL)
170, 145 4.5, 3.9 QN (>10 3
XFU/mL)
83, 161 1.6, 1.7 SQ (>15 CFU) NA
[Local or Sx] or
[CX, NoSource]
CX, Sx
193, 181 5.3, 6. SQ (>15 CFU) CX, Molec,
NoSource
* CFU indicates colony forming units; CX, same organism or species matched between blood
and catheter segment culture; ICU: intensive care units; Local: local signs of infection; Molec:
same organism confirmed by molecular subtyping; NA:not available; NoSource: no other
source of infection; QN: quantitative; Sx: clinical symptoms of bloodstream infection; SQ:
semiquantitative.
† All studies used 10% povidone-iodine solution.
‡ Catheter segments (or site) cultured and criteria for a positive culture are given in parenthesis.
§ Biseptine® consists of 0.25% chlorhexidine gluconate, 0.025% benzalkonium chloride, 4%
benzyl alcohol.
Required one of the following symptoms: fever, erythema, heat at the site, and pain.

Table 16.3.2. Results of Studies Comparing Chlorhexidine Gluconate (CHG) and Povidoneiodine
(PI) Solutions for Vascular Catheter Site Care *
Maki 48
Sheehan 9
Meffre 50
Mimoz 51
Cobett and
LeBlanc 52 †
Humar 53
Catheter Colonization
(Positive Cultures)
CHG PI Solution
Solution
RR (95% CI)
CHG vs. PI
178
Catheter Related
Bloodstream Infection
CHG PI Solution
Solution
RR (95% CI)
CHG vs. PI
5/214 21/227 0.25 (0.10,0.66) 1/214 6/227 0.18 (0.02,1.46)
3/169 12/177 0.22 (0.06,0.75) 1/169 1/177 1.05 (0.07,16.61)
9/568 22/549 0.40 (0.18,0.85) 3/568 3/549 0.97 (0.20,4.77)
12/170 24/145 0.43 (0.22,0.82) 3/170 4/145 0.64 (0.15,2.81)
6/83 23/161 0.49 (0.31,0.77) - - -
36/116 27/116 1.33 (0.87,2.04) 4/193 5/181 0.75 (0.20,2.75)
* CI indicates confidence interval; RR, relative risk.
† Additional information was provided by authors
Subchapter 16.4. Other Practices
Practices That Appear Promising
Use of heparin with central venous catheters. Because an association has been shown between
thrombus formation and catheter-related infection, clinicians usually use heparin, in a variety of
forms: 1) as flushes to fill the catheter lumens between use; 2) injected subcutaneously; or 3)
bonded on the catheter. A meta-analysis of 12 randomized trials evaluating prophylactic use of
heparin in patients using central venous catheters has shown that prophylactic heparin decreases
catheter-related venous thrombosis (Level 2 outcome; RR 0.43, 95% CI: 0.23-078) and bacterial
colonization (Level 2 outcome; RR 0.18, 95% CI: 0.06-0.60) and may decrease CR-BSI (Level 1
outcome; RR 0.26, 95% CI: 0.07-1.03). 59 Since subcutaneous heparin also offers benefit in
reducing venous thromboembolism in certain patient populations (see Chapter 31), this is likely
to be a reasonable strategy even though CR-BSIs have not definitely been shown to be reduced.
However use of heparin is associated with several side effects, such as heparin-induced
thrombocytopenia and bleeding.
Tunneling short-term central venous catheters. Since the primary site of entry for
microorganisms on the central venous catheter is the site of cutaneous insertion, 60 tunneling the
catheter through the subcutaneous tissue may decease the incidence of infection. Several trials
have evaluated the effect of tunneling on catheter-related infection. A recent meta-analysis has
summarized the potential benefit. 61 The meta-analysis included 7 trials and found that compared
with patients receiving standard catheter placement, tunneling decreased bacterial colonization
(Level 2 outcome; RR 0.61, 95% CI: 0.39-0.95) and decreased CR-BSI (Level 1 outcome; RR
0.56, 95% CI: 0.31-1). 61 However, the benefit of tunneling came primarily from one trial using
the internal jugular as the site of catheter placement; the reduction in CR-BSI no longer reached
statistical significance when data from the several subclavian catheter trials were pooled (RR
0.71; 95% CI 0.36-1.43). 61 The authors concluded appropriately that current evidence does not

support the routine use of tunneling central venous catheters. This could change if the efficacy of
tunneling is clearly demonstrated at different placement sites and relative to other interventions
(eg, antiseptic coated catheters). 61
Ineffective Practices
Intravenous antimicrobial prophylaxis. There is no evidence to support the systemic use of
either vancomycin 62 or teicoplanin 63 during insertion of central venous catheters. The
randomized studies evaluating the use on intravenous vanomycin or teicoplanin have failed to
demonstrate that this intervention reduces CR-BSI (Level 1 outcome). 62, 63 Given the theoretical
risk of developing resistance to the antimicrobial agents used for prophylaxis, this practice is not
recommended.
References
1. Raad II, Bodey GP. Infectious complications of indwelling vascular catheters. Clin Infect
Dis 1992;15:197-208.
2. Weinstein MP, Towns ML, Quartey SM, et al. The clinical significance of positive blood
cultures in the 1990s: a prospective comprehensive evaluation of the microbiology,
epidemiology, and outcome of bacteremia and fungemia in adults. Clin Infect Dis
1997;24:584-602.
3. Maki DG. Infections caused by intravascular devices used for infusion therapy. In: Bistro
AI, F.A. W, eds. Infections Associated with Indwelling Medical Devices. Washington, DC:
ASM Press, 1994:155-205.
4. Mermel LA. Prevention of intravascular catheter-related infections. Ann Intern Med 2000;
132:391-402.
5. Guideline for prevention of intravascular device-related infections. Am J Infect Control
1996; 24:261-293.
6. Maki DG, Weise CE, Sarafin HW. A semiquantitative culture method for identifying
intravenous-catheter-related infection. N Engl J Med 1977;296:1305-1309.
7. Sherertz RJ, Heard SO, Raad, II. Diagnosis of triple-lumen catheter infection: comparison
of roll plate, sonication, and flushing methodologies. J Clin Microbiol 1997;35:641-646.
8. Flynn PM, Shenep JL, Barrett FF. Differential quantitation with a commercial blood
culture tube for diagnosis of catheter-related infection. J Clin Microbiol 1988;26:1045-
1046.
9. Blot F, Nitenberg G, Chachaty E, et al. Diagnosis of catheter-related bacteraemia: a
prospective comparison of the time to positivity of hub-blood versus peripheral-blood
cultures. Lancet 1999;354:1071-1077.
10. Bozzetti F, Terno G, Camerini E, Baticci F, Scarpa D, Pupa A. Pathogenesis and
predictability of central venous catheter sepsis. Surgery 1982;91:383-389.
11. Hampton AA, Sherertz RJ. Vascular-access infections in hospitalized patients. Surg Clin
North Am 1988;68:57-71.
12. Corona ML, Peters SG, Narr BJ, Thompson RL. Infections related to central venous
catheters . Mayo Clin Proc 1990;65:979-986.
13. Snydman DR, Gorbea HF, Pober BR, Majka JA, Murray SA, Perry LK. Predictive value of
surveillance skin cultures in total-parenteral-nutrition-related infection. Lancet
1982;2:1385-1388.
179

14. Sitges Serra A, Puig P, Linares J, et al. Hub colonization as the initial step in an outbreak
of catheter-related sepsis due to coagulase negative staphylococci during parenteral
nutrition. JPEN J Parenter Enteral Nutr 1984;8:668-672.
15. Kovacevich DS, Faubion WC, Bender JM, Schaberg DR, Wesley JR. Association of
parenteral nutrition catheter sepsis with urinary tract infections. JPEN J Parenter Enteral
Nutr 1986;10:639-641.
16. Maki DG, Rhame FS, Mackel DC, Bennett JV. Nationwide epidemic of septicemia caused
by contaminated intravenous products. I. Epidemiologic and clinical features. Am J Med
1976;60:471-485.
17. Saint S, Veentra DL, Lipsky BA. The clinical and economic consequences of nosocomial
central venous catheter-related infection: are antimicrobial catheters useful? Infect Control
Hosp Epidemiol 2000;21:375-380.
18. Anonymous. National Nosocomial Infections Surveillance (NNIS) System report, data
summary from October 1986-April 1998, issued June 1998. Am J Infect Control
1998;26:522-533.
19. Pittet D, Tarara D, Wenzel RP. Nosocomial bloodstream infection in critically ill patients.
Excess length of stay, extra costs, and attributable mortality. JAMA 1994;271:1598-601.
20. Pittet D, Wenzel RP. Nosocomial bloodstream infections in the critically ill [letter]. JAMA
1994;272:1819-1820.
21. Smith RL, Meixler SM, Simberkoff MS. Excess mortality in critically ill patients with
nosocomial bloodstream infections. Chest 1991;100:164-167.
22. Digiovine B, Chenoweth C, Watts C, Higgins M. The attributable mortality and costs of
primary nosocomial bloodstream infections in the intensive care unit. Am J Respir Crit
Care Med 1999;160:976-981.
23. Soufir L, Timsit JF, Mahe C, Carlet J, Regnier B, Chevret S. Attributable morbidity and
mortality of catheter-related septicemia in critically ill patients: a matched, risk-adjusted,
cohort study. Infect Control Hosp Epidemiol 1999;20:396-401.
24. Cook D, Randolph A, Kernerman P, et al. Central venous catheter replacement strategies: a
systematic review of the literature. Crit Care Med 1997;25:1417-1424.
25. Eyer S, Brummitt C, Crossley K, Siegel R, Cerra F. Catheter-related sepsis: prospective,
randomized study of three methods of long-term catheter maintenance. Crit Care Med
1990;18:1073-1079.
26. Cobb DK, High KP, Sawyer RG, et al. A controlled trial of scheduled replacement of
central venous and pulmonary-artery catheters . N Engl J Med 1992;327:1062-8.
27. Nystrom B. Impact of handwashing on mortality in intensive care: examination of the
evidence. Infect Control Hosp Epidemiol 1994;15:435-436.
28. Doebbeling BN, Stanley GL, Sheetz CT, et al. Comparative efficacy of alternative handwashing
agents in reducing nosocomial infections in intensive care units. N Engl J Med
1992;327:88-93.
29. Raad, II, Hohn DC, Gilbreath BJ, et al. Prevention of central venous catheter-related
infections by using maximal sterile barrier precautions during insertion. Infect Control
Hosp Epidemiol 1994;15:231-238.
30. Sherertz RJ, Ely EW, Westbrook DM, et al. Education of physicians-in-training can
decrease the risk for vascular catheter infection. Ann Intern Med 2000;132:641-648.
31. Tennenberg S, Lieser M, McCurdy B, et al. A prospective randomized trial of an antibioticand
antiseptic-coated central venous catheter in the prevention of catheter-related
infections. Arch Surg 1997;132:1348-1351.
180

32. Maki DG, Stolz SM, Wheeler S, Mermel LA. Prevention of central venous catheter-related
bloodstream infection by use of an antiseptic-impregnated catheter. A randomized,
controlled trial. Ann Intern Med 1997;127:257-266.
33. van Heerden PV, Webb SA, Fong S, Golledge CL, Roberts BL, Thompson WR. Central
venous catheters revisited–infection rates and an assessment of the new Fibrin Analysing
System brush. Anaesth Intensive Care 1996;24:330-333.
34. Hannan M, Juste R, Shankar U, Nightingale C, Azadian B, Soni N. Colonization of triple
lumen catheters. A study on antiseptic-bonded and standard catheters [abstract]. Clin
Intensive Care 1996;7:56.
35. Bach A, Bohrer H, Bottiger BW, Motsch J, Martin E. Reduction of bacterial colonization
of triple-lumen catheters with antiseptic bonding in septic patients [abstract].
Anesthesiology 1994;81:A261.
36. Bach A, Schmidt H, Bottiger B, et al. Retention of antibacterial activity and bacterial
colonization of antiseptic-bonded central venous catheters. J Antimicrob Chemother
1996;37:315-322.
37. Heard SO, Wagle M, Vijayakumar E, et al. Influence of triple-lumen central venous
catheters coated with chlorhexidine and silver sulfadiazine on the incidence of catheterrelated
bacteremia. Arch Intern Med 1998;158:81-87.
38. Collin GR. Decreasing catheter colonization through the use of an antiseptic-impregnated
catheter: a continuous quality improvement project. Chest 1999;115:1632-1640.
39. Ciresi DL, Albrecht RM, Volkers PA, Scholten DJ. Failure of antiseptic bonding to prevent
central venous catheter-related infection and sepsis. Am Surg 1996;62:641-646.
40. Pemberton LB, Ross V, Cuddy P, Kremer H, Fessler T, McGurk E. No difference in
catheter sepsis between standard and antiseptic central venous catheters. A prospective
randomized trial. Arch Surg 1996;131:986-989.
41. Ramsay J, Nolte F, Schwarmann S. Incidence of catheter colonization and catheter-related
infection with an antiseptic-impregnated triple lumen catheter [abstract]. Crit Care Med
1994;22:A115.
42. Trazzera S, Stern G, Rakesh B, Sinha S, Reiser P. Examination of antimicrobial-coated
central venous catheters in patients at high risk for catheter-related infections in a medical
intensive care unit and leukemia/bone marrow transplant unit [abstract]. Crit Care Med
1995;23:A152.
43. George SJ, Vuddamalay P, Boscoe MJ. Antiseptic-impregnated central venous catheters
reduce the incidence of bacterial colonization and associated infection in
immunocompromised transplant patients. Eur J Anaesthesiol 1997;14:428-431.
44. Veenstra DL, Saint S, Saha S, Lumley T, Sullivan SD. Efficacy of antiseptic-impregnated
central venous catheters in preventing catheter-related bloodstream infection: a metaanalysis.
JAMA 1999;281:261-267.
45. Raad I, Darouiche R, Dupuis J, et al. Central venous catheters coated with minocycline and
rifampin for the prevention of catheter-related colonization and bloodstream infections. A
randomized, double-blind trial. The Texas Medical Center Catheter Study Group. Ann
Intern Med 1997;127:267-274.
46. Darouiche RO, Raad, II, Heard SO, et al. A comparison of two antimicrobial-impregnated
central venous catheters. Catheter Study Group. N Engl J Med 1999;340:1-8.
47. Veenstra DL, Saint S, Sullivan SD. Cost-effectiveness of antiseptic-impregnated central
venous catheters for the prevention of catheter-related bloodstream infection. JAMA
1999;282:554-560.
181

48. Maki DG, Ringer M, Alvarado CJ. Prospective randomised trial of povidone-iodine,
alcohol, and chlorhexidine for prevention of infection associated with central venous and
arterial catheters. Lancet 1991;338:339-343.
49. Sheehan G, Leicht K, O'Brien M, Taylor G, Rennie R. Chlorhexidine versus povidineiodine
as cutaneous antisepsis for prevention of vascular-catheter infection [abstract].
Program and abstracts of the Interscience Conference on Antimicrobial Agents and
Chemotherapy 1993;33:A1616.
50. Meffre C, Girard R, Hajjar J, Fabry J. Is peripheral venous catheter colonisation related to
antiseptic used for disinfection of the insertion site? Hygiene 1995;9:45.
51. Mimoz O, Pieroni L, Lawrence C, et al. Prospective, randomized trial of two antiseptic
solutions for prevention of central venous or arterial catheter colonization and infection in
intensive care unit patients. Crit Care Med 1996;24:1818-1823.
52. Cobett S, LeBlanc A. IV site infection: a prospective, randomized clinical trial comparing
the efficacy of three methods of skin antisepsis: CINA conference '99. CINA: Official
Journal of the Canadian Intravenous Nurses Association 1999;15:48-49.
53. Humar A, Ostromecki A, Direnfeld J, et al. Prospective randomized trial of 10% povidoneiodine
versus 0.5% tincture of chlorhexidine as cutaneous antisepsis for prevention of
central venous catheter infection. Clin Infect Dis 2000;31:1001-1007.
54. Yusuf S. Obtaining medically meaningful answers from an overview of randomized
clinical trials. Stat Med 1987;6:281-294.
55. Light RJ. Accumulating evidence from independent studies: what we can win and what we
can lose. Stat Med 1987;6:221-231.
56. Hennekens CH, Buring JE, Hebert PR. Implications of overviews of randomized trials. Stat
Med 1987;6:397-409.
57. Sacks HS, Berrier J, Reitman D, Ancona Berk VA, Chalmers TC. Meta-analyses of
randomized controlled trials. N Engl J Med 1987;316:450-455.
58. Gelber RD, Goldhirsch A. Interpretation of results from subset analyses within overviews
of randomized clinical trials. Stat Med 1987;6:371-388.
59. Randolph AG, Cook DJ, Gonzales CA, Andrew M. Benefit of heparin in central venous
and pulmonary artery catheters: a meta-analysis of randomized controlled trials. Chest
1998;113:165-171.
60. Mermel LA, McCormick RD, Springman SR, Maki DG. The pathogenesis and
epidemiology of catheter-related infection with pulmonary artery Swan-Ganz catheters: a
prospective study utilizing molecular subtyping. Am J Med 1991;91:197S-205S.
61. Randolph AG, Cook DJ, Gonzales CA, Brun-Buisson C. Tunneling short-term central
venous catheters to prevent catheter-related infection: a meta-analysis of randomized,
controlled trials. Crit Care Med 1998;26:1452-1457.
62. McKee R, Dunsmuir R, Whitby M, Garden OJ. Does antibiotic prophylaxis at the time of
catheter insertion reduce the incidence of catheter-related sepsis in intravenous nutrition? J
Hosp Infect 1985;6:419-425.
63. Ljungman P, Hagglund H, Bjorkstrand B, Lonnqvist B, Ringden O. Peroperative
teicoplanin for prevention of gram-positive infections in neutropenic patients with
indwelling central venous catheters: a randomized, controlled study. Support Care Cancer
1997;5:485-488.
182

Final Comment to Chapter 16
Infections due to central venous catheters are common and lead to substantial morbidity
and health care costs. Several practices will likely reduce the incidence of this common patient
safety problem, including the use of maximum sterile barrier precautions during catheter
insertion, use of catheters coated with antibacterial or antiseptic agents, and use of chlorhexidine
gluconate at the insertion site. Additionally, use of heparin and tunneling of the central venous
catheter may prove to be effective in reducing CR-BSI. However, the relative efficacy of these
interventions is unclear. Also, a clear and formal delineation of the economic consequences of
combining several of these patient safety practices is necessary.
183

184

Chapter 17. Prevention of Ventilator-Associated Pneumonia
Harold R Collard, MD
University of Colorado Health Sciences Center
Sanjay Saint, MD, MPH
University of Michigan School of Medicine
Introduction
Ventilator-associated pneumonia (VAP) is a leading cause of morbidity and mortality in
the intensive care unit (ICU). 1 The incidence of VAP varies greatly, ranging from 6 to 52% of
intubated patients depending on patient risk factors. The cumulative incidence is approximately
1-3% per day of intubation. Overall, VAP is associated with an attributable mortality of up to
30%. Attributable mortality approaches 50% when VAP is caused by the more virulent
organisms that typify late-onset VAP (occurring 4 or more days into mechanical ventilation).
The cost per episode of VAP is substantial, although specific data are lacking. The average cost
per episode of nosocomial pneumonia is estimated at $3000 to $6000, and the additional length
of stay for patients who develop VAP is estimated at 13 days. 1,2
VAP is typically categorized as either early-onset VAP (occurring in the first 3-4 days of
mechanical ventilation) or late-onset VAP. This distinction is important microbiologically.
Early-onset VAP is commonly caused by antibiotic-sensitive community-acquired organisms
(eg, Streptococcus pneumoniae, Haemophilus influenzae, and Staphylococcus aureus). Lateonset
VAP is commonly caused by antibiotic-resistant nosocomial organisms (eg, Pseudomonas
aeruginosa, methicillin-resistant Staphylococcus aureus, Acinetobacter species, and
Enterobacter species). Most episodes of ventilator-associated pneumonia (VAP) are thought to
develop from the aspiration of oropharyngeal secretions containing potentially pathogenic
organisms. Aspiration of gastric secretions may also contribute, though likely to a lesser degree.
Tracheal intubation interrupts the body’s anatomic and physiologic defenses against aspiration,
making mechanical ventilation a major risk factor for VAP.
This chapter reviews 4 practices that carry the potential to reduce the incidence of VAP
in patients receiving mechanical ventilation. They are: variation in patient positioning,
continuous aspiration of subglottic secretions, selective digestive tract decontamination, and the
use of sucralfate.
Subchapter 17.1. Patient Positioning: Semi-recumbent Positioning and Continuous
Oscillation
Background
Aspiration of gastric secretions likely contributes to the development of VAP. 1 Semirecumbent
positioning of mechanically ventilated patients may help reduce the incidence of
gastroesophogeal reflux and lead to a decreased incidence of VAP. Immobility in critically ill
patients leads to atelectasis and decreased clearance of bronchopulmonary secretions. Both of
these sequelae may lead to increased risk of VAP. Continuous rotation and movement of
critically ill patients (termed continuous oscillation) may thus help prevent such changes.
185

Semi-recumbent positioning
Practice Description
Semi-recumbent positioning is generally defined as elevation of the head of the bed to 45
degrees. This is generally achieved in a hospital bed with patients’ feet remaining parallel to the
floor (ie, the entire bed is not tilted) but this is not explicitly described in the published trials.
Semi-recumbency is generally continued for the duration of mechanical ventilation.
Opportunities for Impact
Outside of select medical centers that have studied this practice, semi-recumbent
positioning has not been widely adopted as the standard of care. Thus, such an intervention
would have enormous opportunity for impact should it prove beneficial.
Study Designs
There have been three trials of semi-recumbent patient positioning and its effect on the
incidence of VAP. 3-5 Two of these studies measured aspiration events using nuclear medicine
techniques, the other was a randomized trial with the primary outcome being VAP. In the one
randomized trial, 86 patients were randomized at the time of intubation to semi-recumbent body
position (45 degrees) or supine body position (0 degrees). 3 All patients received the same general
critical care (eg, sterile endotracheal suctioning, stress ulcer prophylaxis with sucralfate if
tolerating oral medications, no ventilator tubing changes, no selective digestive tract
decontamination).
Study Outcomes
In the one randomized clinical trial, VAP was clinically defined as a new and persistent
infiltrate on chest radiography, plus two of the following: temperature of >38.3C, leukocyte
count >12,000/mm 3 or

Potential for Harm
No adverse effects were observed in patients randomized to semi-recumbent positioning. 3
However, patients were excluded if they had any of the following conditions: recent abdominal
or neurologic surgery (

Potential for Harm
There were no significant risks of continuous oscillation in any of the randomized trials.
Inadvertent disconnection of intravenous lines, increased ventricular ectopy, and patient
intolerance were reported, but not quantified. Conscious patients tolerated the procedure poorly.
Costs and Implementation
The incremental cost of specialized beds capable of continuous oscillation has been
estimated at approximately $100 per day. 9 A significant reduction in VAP incidence and length
of stay could result in cost savings.
Comment
Both semi-recumbent positioning and continuous oscillation are relatively low-cost, lowrisk
interventions. The one randomized trial to date of semi-recumbent positioning shows it to be
an effective method of reducing VAP. While it has not proven to provide a mortality benefit,
semi-recumbent positioning is a safe and straightforward intervention whose effectiveness
should be confirmed by additional randomized clinical trials. Continuous oscillation is less
clearly beneficial, although it may be effective in certain subgroups of patients (eg, surgical,
neurologic). It also deserves continued study.
188

Table 17.1.1. Patient positioning*
Study Design Study
Design,
Semi-recumbent positioning
Randomized controlled trial of semi-recumbent
patient positioning in 86 mechanically ventilated
patients. Primary outcome was VAP. (Drakulovic,
1999) 3
Two-period crossover trial of semi-recumbent
patient positioning in 15 mechanically ventilated
patients. Primary outcome was pulmonary
aspiration. (Orozco-Levi, 1995) 4
Randomized two-period crossover trial of semirecumbent
patient positioning in 15 mechanically
ventilated patients. Primary outcome was
pulmonary aspiration. (Torres, 1992) 5
Continuous oscillation
Randomized controlled trial of continuous
oscillation in 103 critically ill medical and surgical
patients (90% mechanically ventilated). Primary
outcomes included pneumonia. (Traver, 1995) 8
Meta-analysis of 6 randomized controlled trials of
continuous oscillation in critically ill surgical or
stroke patients (majority mechanically ventilated).
(Choi, 1992) 7
Randomized controlled trial of continuous
oscillation in 86 critically ill medical patients
(majority mechanically ventilated). Primary
outcomes included pneumonia. (Summer, 1989) 9
189
Outcomes
Level 1,
Level 1
Level 3,
Level 2
Level 3,
Level 2
Level 1,
Level 1
Level 1A,
Level 1
Level 1,
Level 1
* RR indicates relative risk; VAP, ventilator-associated pneumonia.
Pneumonia
or
Aspiration
RR 0.24
(p=0.003)
RR 0.65
(p0.05)
No significant
difference
(data not
reported)
RR 0.93
(p>0.05)

References
1. Craven DE, Steger KA. Nosocomial pneumonia in mechanically ventilated adult patients:
epidemiology and prevention in 1996. Semin Respir Infect. 1996;11:32-53.
2. Thompson R. Prevention of nosocomial pneumonia. Med Clin North Am. 1994;78:1185-
1198.
3. Drakulovic MB, Torres A, Bauer TT, Nicolas JM, Nogue S, Ferrer M. Supine body
position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a
randomised trial. Lancet. 1999;354:1851-1858.
4. Orozco-Levi M, Torres A, Ferrer M, Piera C, el-Ebiary M, de la Bellacasa JP, et al.
Semirecumbent position protects from pulmonary aspiration but not completely from
gastroesophageal reflux in mechanically ventilated patients. Am J Respir Crit Care Med.
1995;152:1387-1390.
5. Torres A, Serra-Batlles J, Ros E, Piera C, Puig de la Bellacasa J, Cobos A, et al. Pulmonary
aspiration of gastric contents in patients receiving mechanical ventilation: the effect of
body position. Ann Intern Med. 1992;116:540-543.
6. Kollef MH. Ventilator-associated pneumonia. A multivariate analysis. JAMA.
1993;270:1965-1970.
7. Choi SC, Nelson LD. Kinetic Therapy in Critically Ill Patients: Combined Results Based
on Meta-Analysis. J Crit Care. 1992;7:57-62.
8. Traver GA, Tyler ML, Hudson LD, Sherrill DL, Quan SF. Continuous oscillation: outcome
in critically ill patients. J Crit Care. 1995;10:97-103.
9. Summer WR, Curry P, Haponik EF, Nelson S, Elston R. Continuous Mechanical Turning
of Intensive Care Unit Patients Shortens Length of Stay in Some Diagnostic-Related
Groups. J Crit Care. 1989;4:45-53.
Subchapter 17.2. Continuous Aspiration of Subglottic Secretions
Background
Ventilator-associated pneumonia (VAP) frequently develops from the aspiration of
oropharyngeal secretions containing potentially pathogenic organisms. 1 Tracheal intubation
interrupts the body’s anatomic and physiologic defenses against aspiration, making mechanical
ventilation a major risk factor for VAP. The accumulation of contaminated oropharyngeal
secretions above the endotracheal tube cuff may contribute to the risk of aspiration. 1 Removal of
these pooled secretions through suctioning of the subglottic region, termed continuous aspiration
of subglottic secretions (CASS), may reduce the risk of developing VAP.
Practice Description
Continuous aspiration of subglottic secretions requires intubation with specially designed
endotracheal tubes (see Figure 17.2.1). These endotracheal tubes contain a separate dorsal lumen
that opens into the subglottic region, allowing for aspiration of any pooled secretions. The
amount of secretions is monitored (usually daily) and the patency of the suction lumen is tested
frequently (every few hours). In studies of the impact of this practice, aspiration has been applied
from time of intubation to time of extubation. One of the studies tested manual aspiration
performed hourly instead of continuous mechanical aspiration. 2
190

Opportunities for Impact
Continuous aspiration of subglottic secretions is an uncommon practice. The
opportunities for impact are therefore significant should this practice prove beneficial in
lowering rates of VAP.
Study Designs
There have been three randomized controlled trials of CASS to date. 2-4 (Table 17.2.1)
Two have included both medical and surgical patients requiring mechanical ventilation for
greater than 72 hours and one included only post-cardiac surgery patients. All three studies
randomized patients to CASS or standard care. Attempts were made to control for additional,
potentially effective preventive strategies such as patient positioning, frequency of ventilator
circuit changes, type of stress ulcer prophylaxis used, and administration of antibiotics.
Study Outcomes
All trials reported development of VAP and mortality at the time of extubation, ICU or
hospital discharge. VAP was generally defined as a new radiographic infiltrate plus two of the
following: fever, leukocytosis/leukopenia, or purulent tracheal aspirate. Microbiologic
confirmation was not consistently obtained. Time to development of VAP was also reported.
Mortality was reported at time of discharge from the hospital.
Evidence for Effectiveness of the Practice
One of the three trials found a statistically significant decrease in the incidence of VAP
with CASS when compared to standard treatment, while a second study showed a strong trend
(See Table 17.2.1). 2,3 All three trials reported a statistically significant delay in the time to
development of VAP, ranging from 48 hours to 8 days. Two trials found a decreased incidence
of VAP caused by Staphylococcus aureus and Hemophilus influenzae, but no change was
observed in the incidence of VAP caused by Pseudomonas aeruginosa or Enterobacteriaceae. 3,4
No difference in mortality was observed in any of the trials.
Potential for Harm
There is minimal potential for harm to patients from the application of CASS and no
adverse patient events were reported in over 150 patients. 4
Costs and Implementation
The cost and cost-effectiveness of CASS have not been examined. The direct costs
appear minimal. Hi-Lo Evac tubes cost approximately 25% more than standard endotracheal
tubes, putting the estimated cost of each unit at less than $1. 2 The cost-savings per episode of
VAP prevented could therefore be substantial. Implementation would largely be a matter of
making the specialized endotracheal tubes available and providing staff training. The mechanical
suctioning apparatus would require frequent monitoring by nursing or respiratory therapy to
insure adequate function.
Comment
Continuous aspiration of subglottic secretions is a promising strategy for the prevention
of VAP. Two randomized controlled trials have suggested a decrease in the rate of VAP in
patients requiring prolonged (>3 days) mechanical ventilation (only one trial was statistically
191

significant). The third trial showed no difference, but the patient population in this trial included
many short-term intubations (mean duration of 36 hours) and was restricted to patients
undergoing cardiac surgery. Larger randomized controlled trials are needed to address the impact
of CASS more definitively.
Another interesting observation is the delay in the development of VAP and the
decreased incidence of Staphylococcus aureus and Hemophilus influenzae. This suggests that
CASS may provide most of its benefit by preventing early VAP caused by community-acquired
organisms, and its use could therefore be targeted to those patients requiring mechanical
ventilation for intermediate periods of time (ie, those at greatest risk for early VAP).
Figure 17.2.1. Diagram of continuous aspiration of subglottic secretions (copied with
permission) 3
192

Table 17.2.1. Randomized trials of continuous aspiration of subglottic secretions*
Study Description Study
Outcomes
Kollef, 1999 4 : 343 patients undergoing
cardiac surgery and requiring mechanical
ventilation
Valles, 1995 3 : 153 patients requiring
prolonged mechanical ventilation
Mahul, 1992 2 : 145 patients requiring
mechanical ventilation for more than 3 days
* CI indicates confidence interval.
193
Relative Risk of
Pneumonia
(95% CI)
Relative Risk
of Mortality
(95% CI)
Level 1 0.61 (0.27-1.40) 0.86 (0.30-2.42)
Level 1 0.47 (0.21-1.06) 1.09 (0.72-1.63)
Level 1 0.46 (0.23-0.93) 1.14 (0.62-2.07)
References
1. Craven DE, Steger KA. Nosocomial pneumonia in mechanically ventilated adult patients:
epidemiology and prevention in 1996. Semin Respir Infect. 1996;11:32-53.
2. Mahul P, Auboyer C, Jospe R, Ros A, Guerin C, el Khouri Z, et al. Prevention of
nosocomial pneumonia in intubated patients: respective role of mechanical subglottic
secretions drainage and stress ulcer prophylaxis. Intensive Care Med. 1992;18:20-25.
3. Valles J, Artigas A, Rello J, Bonsoms N, Fontanals D, Blanch L, et al. Continuous
aspiration of subglottic secretions in preventing ventilator- associated pneumonia. Ann
Intern Med. 1995;122:179-186.
4. Kollef MH, Skubas NJ, Sundt TM. A randomized clinical trial of continuous aspiration of
subglottic secretions in cardiac surgery patients. Chest. 1999;116:1339-1346.
Subchapter 17.3. Selective Digestive Tract Decontamination
Background
Selective digestive tract decontamination (SDD) involves the use of non-absorbable
antibiotics topically applied to the gastrointestinal tract in an effort to sterilize the oropharynx
and stomach. The goal is to decrease the pathogenicity of aspirated secretions and thereby reduce
the incidence of VAP.
Practice Description
Most studies have used a combination of topical polymixin, tobramycin or gentamicin,
and amphotericin applied to the oropharynx (by hand) and the stomach (by nasogastric tube). 1
About half of the studies also included a short (3-4 day) course of systemic intravenous
antimicrobial therapy, most commonly ceftriaxone. In general, topical antibiotics were applied
several times daily from the time of intubation until extubation (or shortly thereafter).

Opportunities for Impact
SDD is not widely used in the United States. 1 The Centers for Disease Control and
Prevention and the American Thoracic Society's guidelines published in the 1990s do not
recommend its routine use. 2,3 Given the frequency and morbidity of VAP, if the practice is
beneficial substantial opportunity for patient safety enhancement exists.
Study Designs
There have been over 30 randomized controlled trials and seven meta-analyses of SDD
(see Table 17.3.1). 4-10 A representative meta-analysis identified 33 randomized trials of SDD
using a structured search of the literature that met the authors’ methodologic inclusion criteria:
measurement of clinical outcomes (including VAP and mortality), inclusion of unselected patient
populations, and mechanical ventilation in at least half of patients. 1 As with several of the other
meta-analyses, individual trials in this particular meta-analysis were grouped into those that used
topical antibiotics only and those that used topical and systemic antibiotics. This meta-analysis
was unique in that the investigators obtained individual patient data for the majority of patients
(4343 (76%) of the 5727 patients involved). 1
Study Outcomes
All meta-analyses reported risk of VAP and mortality at hospital or ICU discharge.
Individual study outcomes also included number of days intubated, length of ICU stay, duration
of antibiotic therapy, time to onset of VAP, and cost. Several meta-analyses performed subgroup
analysis to assess the importance of statistical methods (eg, quality of randomization, blinding,
VAP definition) and clinical factors (eg, Acute Physiology and Chronic Health Evaluation
(APACHE) score).
Evidence for Effectiveness of the Practice
All seven meta-analyses report substantial reduction in the risk of VAP with the use of
SDD (see Table 17.3.1). Four of seven meta-analyses report a statistically significant reduction
in mortality. 4-6,8 Four of seven meta-analyses separately analyzed trials using topical antibiotics
only and those using topical and systemic antibiotics. 4,6,8,9 All four revealed a statistically
significant mortality benefit with combined topical and systemic prophylaxis and no mortality
benefit with topical prophylaxis alone. 1,6,8,9 However, these four meta-analyses did reveal a
significant decrease in VAP incidence in those given topical antibiotics only compared to the
placebo group. 4,6,8,9 Several of the meta-analyses included subgroup analyses to assess the
benefit of SDD in patients categorized by type of illness (surgical, medical) and severity of
illness (APACHE score), with conflicting results. 4,6
Potential for Harm
There were no significant adverse events reported in most trials, although allergic
reactions to the antibiotic preparations have been uncommonly noted. The primary long-term
concern with the widespread use of SDD is the development of antibiotic resistance. 11-13 The
data are unclear regarding the impact of SDD on the emergence of resistant organisms, and no
study has demonstrated an impact of increased bacterial resistance on morbidity or mortality.
Costs and Implementation
The cost of implementing SDD appears minimal in most trials, but there have been no in
depth reviews of the subject. Several trials have found that patients receiving SDD had lower
194

total antibiotic costs. 14-16 Overall hospital costs also may be lower, mediated through the
decreased rate of VAP. 17
Comment
SDD is a very promising method of reducing VAP and ICU-related mortality. The data
supporting a significant reduction in risk of VAP and short-term mortality with SDD using
topical and short-term intravenous antibiotics are strong. SDD is a relatively non-invasive
intervention and the additional financial cost is minimal. What remains to be determined is the
long-term effect of SDD on antibiotic resistance patterns, and the impact of such effect on
morbidity and mortality. Research into the impact of SDD on the emergence of antibiotic
resistance should be strongly encouraged.
195

Table 17.3.1. Meta-analyses of selective digestive tract decontamination*
Study Design Pneumonia (95% CI) Mortality (95% CI)
Nathens, 1999 6 : 21 randomized
controlled trials of antibiotic
prophylaxis used to decrease
nosocomial respiratory tract infections;
dual analysis of medical and surgical
patients
D’Amico, 1998 1 : 33 randomized
controlled trials from of antibiotic
prophylaxis used to decrease
nosocomial respiratory tract infections;
dual analysis of topical and systemic
antibiotics combined and topical
antibiotics alone
Hurley, 1995 5 : 26 randomized
controlled trials of antibiotic
prophylaxis used to decrease
nosocomial respiratory tract infections
Kollef, 1994 7 : 16 randomized
controlled trials of antibiotic
prophylaxis used to decrease
nosocomial respiratory tract infections
Heyland, 1994 8 : 25 randomized
controlled trials of antibiotic
prophylaxis used to decrease
nosocomial respiratory tract infections;
performed subgroup analyses
SDD Trialists’ Collaborative Group,
1993 9 : 22 randomized controlled trials
of antibiotic prophylaxis used to
decrease nosocomial respiratory tract
infections; performed subgroup
analyses
Vandenbroucke-Grauls, 1991 10 : 6
randomized controlled trials of
antibiotic prophylaxis used to decrease
nosocomial respiratory tract infections
Medical: OR 0.45 (0.33-
0.62)
Surgical: OR 0.19 (0.15-
0.26)
Overall: not reported
Topical/IV: OR 0.35
(0.29-0.41)
Topical: OR 0.56 (0.46-
0.68)
Overall: OR 0.35 (0.30-
0.42)
Overall: RD 0.145
(0.116-0.174)
Overall: RR 0.46 (0.39-
0.56)
Topical/IV: RR 0.48
(0.39-0.60)
Topical: RR 0.43 (0.32-
0.59)
Overall: OR 0.37 (0.31-
0.43)
Topical/IV: OR 0.33
(0.27-0.40)
Topical: OR 0.43 (0.33-
0.56)
Overall: OR 0.12 (0.08-
0.19)
196
Medical
Overall: OR 0.91 (0.71-1.18)
Topical/IV: OR 0.75 (0.53-1.06)
Topical: OR 1.14 (0.77-1.68)
Surgical
Overall: OR 0.70 (0.52-0.93)
Topical/IV: OR 0.60 (0.41-0.88)
Topical: OR 0.86 (0.51-1.45)
Overall: OR 0.88 (0.78-0.98)
Topical/IV: OR 0.80 (0.69-0.93)
Topical: OR 1.01 (0.84-1.22)
Overall: OR 0.86 (0.74-0.99)
Overall: RD 0.019 (-0.016-0.054)
Overall: RR 0.87 (0.79-0.97)
Topical/IV: RR 0.81 (0.71-0.95)
Topical: RR 1.00 (0.83-1.19)
Overall: OR 0.90 (0.79-1.04)
Topical/IV: OR 0.80 (0.67-0.97)
Topical: OR 1.07 (0.86-1.32)
Overall: OR 0.70 (0.45-1.09)
* CI indicates confidence interval; RD, risk difference, RR, relative risk; and OR, odds ratio.

References
1. D'Amico R, Pifferi S, Leonetti C, Torri V, Tinazzi A, Liberati A. Effectiveness of
antibiotic prophylaxis in critically ill adult patients: systematic review of randomised
controlled trials. BMJ. 1998;316:1275-1285.
2. Hospital-acquired pneumonia in adults: diagnosis, assessment of severity, initial
antimicrobial therapy, and preventive strategies. A consensus statement, American
Thoracic Society, November 1995. Am J Respir Crit Care Med. 1996;153:1711-1725.
3. CDC. Guidelines for prevention of nosocomial pneumonia. Centers for Disease Control
and Prevention. MMWR Morb Mortal Wkly Rep. 1997;46:1-79.
4. Craven DE, Steger KA. Nosocomial pneumonia in mechanically ventilated adult patients:
epidemiology and prevention in 1996. Semin Respir Infect. 1996;11:32-53.
5. Hurley JC. Prophylaxis with enteral antibiotics in ventilated patients: selective
decontamination or selective cross-infection? Antimicrob Agents Chemother. 1995;39:941-
947.
6. Nathens AB, Marshall JC. Selective decontamination of the digestive tract in surgical
patients: a systematic review of the evidence. Arch Surg. 1999;134:170-176.
7. Kollef MH. The role of selective digestive tract decontamination on mortality and
respiratory tract infections. A meta-analysis. Chest. 1994;105:1101-1108.
8. Heyland DK, Cook DJ, Jaeschke R, Griffith L, Lee HN, Guyatt GH. Selective
decontamination of the digestive tract. An overview. Chest. 1994;105:1221-1229.
9. Meta-analysis of randomised controlled trials of selective decontamination of the digestive
tract. Selective Decontamination of the Digestive Tract Trialists' Collaborative Group.
BMJ. 1993;307:525-532.
10. Vandenbroucke-Grauls CM, Vandenbroucke JP. Effect of selective decontamination of the
digestive tract on respiratory tract infections and mortality in the intensive care unit.
Lancet. 1991;338:859-862.
11. van Saene HK, Stoutenbeek CP, Hart CA. Selective decontamination of the digestive tract
(SDD) in intensive care patients: a critical evaluation of the clinical, bacteriological and
epidemiological benefits. J Hosp Infect. 1991;18:261-277.
12. Ebner W, Kropec-Hubner A, Daschner FD. Bacterial resistance and overgrowth due to
selective decontamination of the digestive tract. Eur J Clin Microbiol Infect Dis.
2000;19:243-247.
13. Bartlett JG. Selective decontamination of the digestive tract and its effect on antimicrobial
resistance. Crit Care Med. 1995;23:613-615.
14. Quinio B, Albanese J, Bues-Charbit M, Viviand X, Martin C. Selective decontamination of
the digestive tract in multiple trauma patients. A prospective double-blind, randomized,
placebo-controlled study. Chest. 1996;109:765-772.
15. Verwaest C, Verhaegen J, Ferdinande P, Schetz M, Van den Berghe G, Verbist L, et al.
Randomized, controlled trial of selective digestive decontamination in 600 mechanically
ventilated patients in a multidisciplinary intensive care unit. Crit Care Med. 1997;25:63-
71.
16. Sanchez Garcia M, Cambronero Galache JA, Lopez Diaz J, Cerda Cerda E, Rubio Blasco
J, Gomez Aguinaga MA, et al. Effectiveness and cost of selective decontamination of the
197

digestive tract in critically ill intubated patients. A randomized, double-blind, placebocontrolled,
multicenter trial. Am J Respir Crit Care Med. 1998;158:908-916.
17. Silvestri L, Mannucci F, van Saene HK. Selective decontamination of the digestive tract: a
life saver. J Hosp Infect. 2000;45:185-190.
Subchapter 17.4. Sucralfate and Prevention of VAP
Background
Aspiration of gastric secretions may contribute to the development of VAP. It has been
observed that gastric colonization by potentially pathogenic organisms increases with decreasing
gastric acidity, leading to the hypothesis that pH-altering drugs may cause increased rates of
VAP. 1 H2-antagonist therapy, widely used in mechanically-ventilated patients for stress ulcer
prophylaxis (see Chapter 34), significantly elevates gastric pH. Sucralfate, an alternative
prophylactic agent that does not affect gastric pH, may allow less gastric colonization with
potentially pathogenic organisms than H2-antagonists and therefore prevent some cases of VAP.
Practice Description
In general, 1 g of sucralfate suspension is given through a nasogastric tube every four to
six hours. When H2-antagonists are used, their dosing and frequency vary. A representative study
used 50 mg of ranitidine intravenously every eight hours, dose adjusted for creatinine clearance. 2
Stress ulcer prophylaxis is usually started upon initiation of mechanical ventilation and
continued until extubation (Chapter 34).
Opportunities for Impact
Stress ulcer prophylaxis is usually given to critically ill ventilated patients. A large cohort
study of over 2000 critically ill patients suggests that the majority receive H2-antagonists
(71.8%) followed by sucralfate (7.0%) and combination therapy (15.4%) or other single agents
(omeprazole, antacids, prostaglandins). 3 There is, therefore, significant opportunity for impact
should sucralfate prove to lower rates of VAP and improve survival.
Study Designs
There have been over 20 randomized controlled trials of stress ulcer prophylaxis using
sucralfate, H2-antagonists, and other therapies in critically ill patients. Seven meta-analyses have
been published to date. 2,4-10 The individual trials and meta-analyses have significant variation in
methodology. In general, the individual trials randomized critically ill patients to sucralfate, H2antagonists,
other agents such as antacids and pirenzepine, or placebo. The majority of patients
included in these studies required mechanical ventilation. Various drug-drug, drug-placebo
combinations were compared, and the rates of VAP and mortality were recorded.
Study Outcomes
Most trials report development of VAP and mortality as primary endpoints. There is
significant variation in the definition of VAP used in these trials. In the largest and most recent
randomized trial, VAP was defined as a new radiographic infiltrate plus two of the following:
temperature of >38.5°C or 10,000/mm 3 or

Evidence for Effectiveness of the Practice
The results of the seven meta-analyses and one recent large randomized controlled trial
are inconclusive (see Table 17.4.1). The two largest meta-analyses to date suggest a decreased
incidence of VAP with sucralfate compared to H2-antagonists, 4,6 and one reports a statistically
significant mortality benefit with sucralfate. 6 A recent randomized controlled trial of 1200
ventilated patients reports no significant difference between the two therapies in terms of VAP or
mortality. 8
Potential for Harm
Sucralfate therapy has been associated with a statistically significant increased risk of
clinically important gastrointestinal bleeding when compared to H2-antagonists. 8 Clinically
important bleeding developed in 3.8% of patients receiving sucralfate compared with 1.7% of
patients receiving H2-antagonists (relative risk for H2-antagonist = 0.44, 95% CI: 0.21-0.92). 8
Gastrointestinal bleeding in critically ill patients has an attributable mortality of approximately
12.5%. 8 While previous meta-analyses have suggested little difference in rates of gastrointestinal
bleeding between the various prophylactic agents, 6 these results from a large randomized trial are
convincing. There are very few adverse effects from sucralfate therapy aside from constipation,
rare nausea and vomiting and very rare bezoar formation and aluminum intoxication. 11 Sucralfate
administration has been associated with transmission of vancomycin resistant enterococcus,
likely due to increased manipulation of patients’ nasogastric tubes. 12 Unlike parenteral H2blockers,
sucralfate mandates nasogastric tube placement in intubated patients. The drug can also
lead to decreased absorption of other medications.
Costs and Implementation
Several studies have looked at the cost-effectiveness of stress ulcer prophylaxis. 13,14
Based on decision analysis, the cost per episode of gastrointestinal bleeding averted in high-risk
patients is several thousand dollars greater with H2-antagonists than with sucralfate. 13 This cost
difference remains significant even if H2-antagonists are assumed to be 50% more effective.
There are no reliable data comparing overall costs from the actual clinical trials. The mean cost,
largely driven by prolonged length of stay, is significantly higher for patients who bleed than for
those who do not ($70,000 vs. $15-20,000), implying that in patients at high risk for GI bleeding
(eg, mechanically ventilated patients, those with a coagulopathy), stress ulcer prophylaxis may
be cost-neutral or even cost-saving (see also Chapter 34). 14 Implementation of sucralfate use
would be largely an issue of staff education as administration is relatively uncomplicated.
Comment
The data supporting stress ulcer prophylaxis with sucralfate instead of H2-antagonists to
prevent VAP are inconclusive, and the theoretical contribution of increased gastric colonization
with potentially pathogenic organisms to the development of VAP is unproven. There are data
both supporting and refuting a decreased incidence of VAP with sucralfate. Most investigators
have found at least a trend toward decreased incidence of VAP with sucralfate, and larger studies
are warranted. The greatest benefit from sucralfate may be the prevention of late-onset VAP in
patients requiring long-term ventilation. 15 Any increased risk of gastrointestinal bleeding with
sucralfate therapy in these patients may be offset by the decreased risk of VAP. Until the data are
more definitive, however, when stress ulcer prophylaxis is deemed appropriate (Chapter 34), the
199

use of H2-blockers seems preferable to sucralfate because of the former’s superiority in
preventing clinically important gastrointestinal bleeding.
200

Table 17.4.1. Studies of stress ulcer prophylaxis
Study Design Design,
Outcomes
Meta-analysis of randomized
controlled trials comparing
ranitidine with placebo,
sucralfate with placebo and
ranitidine with sucralfate for the
prevention of pneumonia in
critically ill patients. (Messori,
2000)
Multicenter randomized, blinded,
placebo-controlled trial of
sucralfate with ranitidine in 1200
critically ill mechanically
ventilated patients. Endpoints
were gastrointestinal bleeding,
VAP and mortality. (Cook,
1998)
27 randomized trials of stress
ulcer prophylaxis in critically ill
patients. The majority of patients
were mechanically ventilated.
Endpoints were gastrointestinal
bleeding, pneumonia and
mortality. (Cook, 1996)
14 randomized trials of stress
ulcer prophylaxis in critically ill
patients. (Tryba, 1995)
6 (outcome VAP) and 7
(outcome mortality) randomized
trials of stress ulcer prophylaxis
in critically ill patients. (Cook,
1995)
14 randomized trials of stress
ulcer prophylaxis in critically ill
patients. Endpoints were
gastrointestinal bleeding and
pneumonia. (Tryba. 1991)
Level 1A,
Level 1
Level 1,
Level 1
Level 1A,
Level 1
Level 1A,
Level 1
Level 1A,
Level 1
Level 1A,
Level 1
201
Pneumonia*
(95% CI)
ranitidine vs.
sucralfate: 1.35
(1.07-1.70)
ranitidine vs.
placebo: 0.98
(0.56-1.72)
sucralfate vs.
placebo: 2.21
(0.86-5.65)
ranitidine vs.
sucralfate: 1.18
(0.92-1.51)
sucralfate vs. H2antagonist:
0.77
(0.60-1.01)
H2-antagonist vs.
placebo: 1.25
(0.78-2.00)
sucralfate vs. H2antagonist/
antacid: 0.67
(p

Table 17.4.1. Studies of stress ulcer prophylaxis (cont.)
9 randomized trials of stress
ulcer prophylaxis in critically ill
patients. (Tryba, 1991)
8 randomized trials of stress
ulcer prophylaxis in critically ill
patients studying the rate of
pneumonia with different drug
regimens. (Cook, 1991)
Level
1A,
Level 1
Level
1A,
Level 1
* Point estimates reflect odds ratio or relative risk.
References
sucralfate vs. H2antagonist/
antacid:
0.48 (p

12. Slaughter S, Hayden MK, Nathan C, Hu TC, Rice T, Van Voorhis J, et al. A comparison of
the effect of universal use of gloves and gowns with that of glove use alone on acquisition
of vancomycin-resistant enterococci in a medical intensive care unit. Ann Intern Med.
1996; 125: 448-456.
13. Ben-Menachem T, McCarthy BD, Fogel R, Schiffman RM, Patel RV, Zarowitz BJ, et al.
Prophylaxis for stress-related gastrointestinal hemorrhage: a cost effectiveness analysis.
Crit Care Med. 1996; 24: 338-345.
14. Erstad BL, Camamo JM, Miller MJ, Webber AM, Fortune J. Impacting cost and
appropriateness of stress ulcer prophylaxis at a university medical center. Crit Care Med.
1997; 25: 1678-1684.
15. Prod'hom G, Leuenberger P, Koerfer J, Blum A, Chiolero R, Schaller MD, et al.
Nosocomial pneumonia in mechanically ventilated patients receiving antacid, ranitidine, or
sucralfate as prophylaxis for stress ulcer. A randomized controlled trial. Ann Intern Med.
1994; 120: 653-662.
Final Comment to Chapter 17
Ventilator-associated pneumonia is common, costly, and morbid. This chapter confirms
that there are several low-risk interventions that carry the potential to reduce the frequency of
this complication. Further research will be needed to confirm the benefit of promising practices
(eg, semi-recumbency or continuous aspiration of subglottic secretions) or fully allay concerns
regarding practices that have potential for harm (eg, antibiotic resistance with selective
decontamination).
203

Section C. Surgery, Anesthesia, and Perioperative Medicine
Chapter 18. Localizing Care to High-Volume Centers
Chapter 19. Learning Curves for New Procedures – the Case of Laparoscopic
Cholecystectomy
Chapter 20. Prevention of Surgical Site Infections
Subchapter 20.1 Prophylactic Antibiotics
Subchapter 20.2. Perioperative Normothermia
Subchapter 20.3. Supplemental Perioperative Oxygen
Subchapter 20.4. Perioperative Glucose Control
Chapter 21. Ultrasound Guidance of Central Vein Catheterization
Chapter 22. The Retained Surgical Sponge
Chapter 23: Pre-Anesthesia Checklists To Improve Patient Safety
Chapter 24. The Impact Of Intraoperative Monitoring On Patient Safety
Chapter 25. Beta-blockers and Reduction of Perioperative Cardiac Events
204

Chapter 18. Localizing Care to High-Volume Centers
Andrew D. Auerbach, MD, MPH
University of California, San Francisco School of Medicine
Background
The extent to which experience in caring for illness—as represented by a higher volume
of cases—impacts outcomes has been well studied over the last 20 years. An extensive literature
covering a broad range of conditions and procedures documents superior outcomes for hospitals
and physicians with higher patient volumes. 1-6 Drawing on such evidence, various investigators
have projected substantial reductions in mortality from regionalizing certain high-risk procedures
with established volume-outcome relationships. 4-6
When volume-outcomes relationships reflect a “practice makes perfect” effect, it may be
reasonable to use volumes to assess quality of care. However, such relationships may also reflect
“selective referral,” 7-10 when anecdotal knowledge of the superior quality of high volume centers
exists in the community. 11 In such cases, direct measurements of processes or outcomes may
represent more appropriate quality measures than volume alone. 12,13
In an era of cost containment and a growing need for accountability for quality of care,
any connection between site or physician-specific experience and patient outcomes has far
reaching implications for patients, payers, and governmental agencies. 14 In fact, the Leapfrog
Group (a consortium of major purchasers and purchasing coalitions) has made patient volume
one of their criteria for quality, and has recently begun a project examining evidence-based
referrals to high-volume centers (see also Chapter 55). 15
The Institute of Medicine (IOM) recently sponsored a workshop to examine the evidence
supporting this relationship, 16 part of which included a systematic review of investigations of the
volume-outcome association. 17 This chapter summarizes the evidence supporting volumeoutcomes
relationships, drawing heavily on the IOM’s systematic review and the workshop's
findings.
Practice Description
The use of information regarding volume and its proven or potential relationship with
better outcomes may result in several actions. Simply providing patients with volume data may
result in preferential selection of high-volume centers or providers. Patients might also be
incentivized to choose high-volume centers (eg, through reduced co-payments). Alternatively,
payers may elect to contract only with high-volume centers, or provide higher payments to these
sites. Finally, low-volume centers (eg, a hospital failing to meet a minimum threshold of bypass
operations) or providers (eg, a cardiologist failing to perform a minimum number of
angioplasties) might be restricted from continuing to perform the practice, through some
combination of credentialing, accreditation, or regulatory actions (see Chapter 56).
Opportunities for Impact
Assuming that a feasible method could be developed to localize care to high-volume
centers, a significant effect on patient safety and outcomes is likely. A recent study suggested
that more than 500 deaths could be avoided annually in California alone if care for disorders
with established volume-outcomes relationships were localized to more experienced centers.
Extrapolating nationally, such localization would save 4000 lives. 6 In the case of acute
myocardial infarction, transferring the care of these patients from hospitals in the lowest volume
205

quartile to those in the highest would save 2.3 lives per 100 patients. 4 The corresponding
“number needed to treat” (NNT) of 50 falls within the range of many accepted therapeutic
interventions.
Little information exists to assess differences in quality of care or rates of adverse events
in high or low-volume sites. For example, the extent to which increasing volume leads to fewer
medical errors or other direct impacts on patient safety (as opposed to specific improvement in
care processes for discrete procedures, which would fall outside our definition of patient safety
practices (Chapter 1)) is unknown. Ongoing prospective initiatives such as those proposed by the
Leapfrog Group 15 may better quantify the various impacts of localizing care to high-volume
centers.
Study Designs
We analyzed one large systematic review of 88 studies examining the relationship
between volume and outcomes. 17 Using a structured MEDLINE search, this review included
studies that examined health outcomes as the dependent variable with hospital and/or physician
volume as an independent variable. The IOM review included medical and surgical conditions
such as coronary artery bypass grafting, pediatric cardiac surgery, carotid endarterectomy,
abdominal aortic aneurysm repair, cancer surgery, coronary angioplasty, acute myocardial
infarction, AIDS, and multiple procedures. This chapter reviews the relationship for surgical
illnesses only.
The source studies of the IOM review were entirely observational in nature. Close
attention was paid to risk adjustment and statistical methods in assessment of results, and criteria
for inclusion in the review selected for population or community-based samples. Thus, the level
of the design is classified as Level 3A.
Study Outcomes
The IOM systematic review examined health outcomes as related to hospital or physician
volume. The primary outcome of interest was mortality. Other clinical outcomes were chosen
based on complications specific to the surgical procedure (eg, stroke following carotid
endarterectomy).
Evidence for Effectiveness of the Practice
Results of the systematic review are outlined in Table 18.1. The studies reviewed were of
variable design and analytic sophistication, with more recent investigations generally being of
higher quality. For all procedures, there was a consistent trend toward an association between
improved outcomes and higher hospital or physician-specific volume.* The evidence supporting
the volume-outcomes relationship was similar when looking at hospital volume (78% of studies
* This trend has one major exception, a study of volume-outcome relationships for 8 major
surgical procedures at Veterans Affairs hospitals across the country (Khuri SF, et al. Relation of
surgical volume to outcome in eight common operations: results from the VA National Surgical
Quality Improvement Program. Ann Surg. 1999;230:414-429). This comprehensive national
study found no significant volume-outcome relationship for any of the 8
procedures analyzed. While it is tempting to attribute these negative findings to unique aspects
of the VA system, this is also one of the few studies to employ robust risk-adjustment using
clinical and not administrative data. The authors of the IOM review take particular note of the
superior methodologic features and negative findings of this study.
206

showed an association) and physician volume (74% showed an association); the former was the
more frequently analyzed variable.
The impact of volume on outcomes varied across procedures and diagnoses. The effect
was most marked for complex cancer surgeries (ie, esophagectomy, pancreatoduodenectomy),
with numbers needed to treat (NNT) between 7 and 17. For commonly performed surgeries such
as coronary artery bypass grafting, the NNT was generally larger, but still within the range of
other accepted therapies. Carotid endarterectomy appeared to have a much higher NNT, but this
may be because the major adverse outcome of this surgery (ie, stroke) is often an indication for
surgery as well as a complication of it. This relationship cannot be discerned from administrative
data, and the studies upon which the NNT is based reported mortality, a less frequent
complication of carotid surgery than stroke, as a primary outcome. The few studies that collected
primary data (and would have been able to determine this important difference) were generally
small and of lower quality design, making their findings suspect.
The authors of the IOM systematic review conclude that the volume-outcomes
relationship exists, but they raise some caveats when interpreting the literature as a whole. They
first note that the literature describing the relationship in greatest detail come from a few single-
State databases (63% of studies coming from New York State alone), possibly limiting the
generalizability of their results. Although few studies employed rigorous risk adjustment
methodologies using clinical data, the relationship between volume and outcomes was consistent
in these studies. Also, they raise a note of caution in interpreting this relationship because of the
possibility that publication bias (ie, studies that fail to demonstrate a volume-outcome
relationship might be less likely to be submitted or accepted for publication) has affected this
literature. Finally, they point out that the precise mechanism by which outcomes are improved
has yet to be elucidated; no study has reported the independent effects of ancillary personnel
expertise or hospital system factors on patient outcomes. For example, in interpreting the
improved outcomes of high volume centers in coronary artery bypass grafting, we do not
presently know what the relative contributions are of the surgeon, cardiac bypass team, cardiac
anesthesiologist, and hospital resources (eg, dedicated cardiothoracic intensive care units).
Potential for Harm
In the summary accompanying the IOM workshop's proceedings, several potential
pitfalls of localizing care to high volume settings were noted, as follows 16 :
• The focus on high volume providers may be a “distracting priority,” and
similar improvements in care may be achieved through more traditional local
quality improvement measures.
• Hospitals with high volumes may use that data to misrepresent their
experience in the absence of true outcomes data.
• High volume centers may achieve too much contractual leverage, leading to
price inflation.
• Counting procedures may lead to perverse incentives to perform procedures
that are not appropriate.
• Requiring high volumes will impede entry of new competitors into the
marketplace.
• Narrowing the choice of providers may negatively impact patient satisfaction
and override patients’ preferences for care (for example, if patients are forced
to travel long distances to receive care at a high volume center).
207

Several of these concerns have been borne out in published studies. In a study of
pediatric trauma centers, Tepas et al suggested that increases in volume may strain provider
resources and worsen patient outcomes. 18 Even assuming potential improvements in patient
outcomes, diverting patients to large referral centers has important health policy implications 14
and may decrease patient satisfaction. 19
Costs and Implementation
The major barriers to implementing a selective referral program based on hospital
volume include the potential harms listed above, as well as several additional factors. These may
include patients' preference for care near home, lack of resources to travel, inability to transfer
unstable patients to high-volume centers, loss of access to care in areas where low-volume
services are discontinued (particularly rural areas), and resistance by providers to quality
measurement activities. Finally, existing high volume centers may lack the capacity to accept
additional patients. When they do not, further increases in volume could lead to increased rates
of adverse events due to over-stressing the system, as was noted for pediatric trauma. 18
Costs of this practice are not explicitly addressed in the IOM report, but widespread
implementation of selective referrals would depend on the collection of detailed and accurate
data (risk adjustment, process, and outcomes data), at substantial cost. Additionally, a
nationwide systematic referral model may require augmenting capability of high-volume
hospitals through additional construction or other major infrastructure investments. Finally, the
travel and inconvenience costs of having patients obtain care at institutions outside the local
community will be borne by either the system or the patients themselves—a cost which may be
compounded further by patients’ and families’ lost productivity.
Comment
Changing practice patterns based on the compelling data linking volume and improved
outcomes is a complex task involving patients, hospitals and communities, as well as payers and
employers. Actually closing hospitals or designating specific health care centers as “magnet”
sites for care of specific illnesses would require, in addition to a wholesale change in health care
systems, major infusions of resources and difficult political choices. For these reasons alone, this
policy decision seems unlikely. Alternatively, hospitals or medical groups may use physicianspecific
outcomes information to make decisions about staffing needs (eg, hiring a lower number
of surgeons to ensure that all operators have high volumes), although this too seems unlikely
unless meeting volume thresholds become mandatory or are strongly incentivized.
The Leapfrog Group's efforts represent one of the first major initiatives to use empiric
data to direct selective referrals using volume data. As an initiative begun and carried out within
a specific employer/health care purchaser system it may be limited in its generalizability.
However, prospective evaluation of the effort will yield important information regarding the
costs and outcomes of such a referral system and its impact on patient satisfaction with care.
Outside of the Leapfrog effort, the widespread publication of outcomes, especially
mortality data, has been proposed as a way to help consumers to make more informed health care
choices. A recent systematic review of this paradigm or “strategy” suggests that its impact has
been mixed. 20 There are no data to suggest that patients or payers will make major changes in
health care purchasing decisions when provided with volume data alone. To date, press reports
of particularly noteworthy errors seem to have more of an impact on patients' choices of care
than information about volume or even outcomes of care. 20,21
208

In addition to the potential use of volume data to guide health care purchasing decisions,
the IOM workshop and authors of the systematic review recommend using volume as one of
several quality measures to initiate local or regional quality improvement efforts. This data may
motivate or inform care improvement efforts at low-volume sites (or those with worse-thanexpected
outcomes) through use of site visits, feedback of risk-adjusted outcomes information,
and assessment of care processes. This collaborative approach to quality improvement has been
used successfully in the VA National Surgical Quality Improvement Project 22 and in several
projects in New York State. 23,24 However, it seems likely that as the volume-outcome
relationship becomes better defined and understood, its influence on the health care choices of
both patients and payers is likely to grow.
209

Table 18.1. Summary of findings from IOM systematic review 17 of volume-outcome
relationship*
Condition Number of
Studies
Reviewed
Coronary
artery bypass
grafting
9
Carotid
endarterectomy 19
Cancer surgery 20
Abdominal
aortic
aneurysm
repair
Pediatric
cardiac
surgery
12
3
Comments
All studies used appropriate risk adjustment
VOR for both physicians and hospital: 7 studies
Absolute difference in mortality between high- and low-volume
centers and surgeons was 3-10% (NNT 10-33)
Some evidence to suggest that centers and surgeons with good
outcomes experienced increasing volumes over time
(“selective referrals”)
Only 9 studies performed adequate risk adjustment
Most studies employed administrative data, making accurate
ascertainment of postoperative stroke impossible
VOR found for surgeon: 9 studies
VOR found for hospital: 7 studies
Absolute difference in mortality between high- and low-volume
hospital/surgeon was 0.2-0.9% (NNT 100-500)
Risk adjustment methods variable, most are dependent on
administrative data only
VOR most marked for rare cancers/procedures such as pancreatic
resection and esophageal surgery
VOR unclear for common surgery such as colectomy and
pneumonectomy
For esophagectomy, absolute difference in mortality between
high- and low-volume hospitals was 11-13.9% (NNT 7-9)
For pancreatic resection, difference in mortality between high- and
low-volume hospitals gives NNT 10-15
11 studies performed adequate risk adjustment
VOR for hospitals and physicians noted
Absolute reduction in mortality due to surgeon or hospital volume
was 5-9% (NNT 11-20)
“Selective referrals” noted
All studies used appropriate risk adjustment
VOR found for both hospital and surgeon volume
Absolute difference in mortality due to surgeon or hospital volume
was 3% (NNT 33)
Possibly greater benefit for more complex/sicker patients
* NNT indicates number needed to treat; VOR, volume-outcome relationship
210

References
1. Luft HS, Bunker JP, Enthoven AC. Should operations be regionalized? The empirical
relation between surgical volume and mortality. N Engl J Med. 1979;301:1364-1369.
2. Begg CB, Cramer LD, Hoskins WJ, Brennan MF. Impact of hospital volume on operative
mortality for major cancer surgery. JAMA. 1998;280:1747-1751.
3. Glasgow RE, Showstack JA, Katz PP, Corvera CU, Warren RS, Mulvihill SJ. The
relationship between hospital volume and outcomes of hepatic resection for hepatocellular
carcinoma. Arch Surg. 1999;134:30-35.
4. Thiemann DR, Coresh J, Oetgen WJ, Powe NR. The association between hospital volume
and survival after acute myocardial infarction in elderly patients. N Engl J Med.
1999;340:1640-1648.
5. Birkmeyer JD, Lucas FL, Wennberg DE. Potential benefits of regionalizing major surgery
in Medicare patients. Eff Clin Pract. 1999;2:277-283.
6. Dudley RA, Johansen KL, Brand R, Rennie DJ, Milstein A. Selective referral to highvolume
hospitals: estimating potentially avoidable deaths. JAMA. 2000;283:1159-1166.
7. Luft HS, Hunt SS, Maerki SC. The volume-outcome relationship: practice-makes-perfect
or selective- referral patterns? Health Serv Res. 1987;22:157-182.
8. Hughes RG, Garnick DW, Luft HS, McPhee SJ, Hunt SS. Hospital volume and patient
outcomes. The case of hip fracture patients. Med Care. 1988;26:1057-1067.
9. Farley DE, Ozminkowski RJ. Volume-outcome relationships and in-hospital mortality: the
effect of changes in volume over time. Med Care. 1992;30:77-94.
10. Hamilton BH, Ho V. Does practice make perfect? Examining the relationship between
hospital surgical volume and outcomes for hip fracture patients in Quebec. Med Care.
1998;36:892-903.
11. Luft HS, Garnick DW, Mark DH, et al. Does quality influence choice of hospital? JAMA.
1990;263:2899-2906.
12. Flood AB, Scott WR, Ewy W. Does practice make perfect? Part I: The relation between
hospital volume and outcomes for selected diagnostic categories. Med Care. 1984;22:98-
114.
13. Luft HS, Hunt SS. Evaluating individual hospital quality through outcome statistics. JAMA.
1986;255:2780-2784.
14. Phillips KA, Luft HS. The policy implications of using hospital and physician volumes as
"indicators" of quality of care in a changing health care environment. Int J Qual Health
Care. 1997;9:341-348.
15. The Leapfrog Group for patient safety: Rewarding higher standards. In: Group TL, ed.
www.leapfrog.org. 2001.
16. Hewitt M. (National Academy of sciences). Interpreting the volume-outcome relationship
in the context of health care quality. 2000.
17. Halm EA, Lee C, Chassin M. (Institute of Medicine). How is volume related to quality in
health care?: A systematic review of the research literature. 2000.
18. Tepas JJ, 3rd, Patel JC, DiScala C, Wears RL, Veldenz HC. Relationship of trauma patient
volume to outcome experience: can a relationship be defined? J Trauma. 1998;44:827-830.
19. Finlayson SR, Birkmeyer JD, Tosteson AN, Nease RF, Jr. Patient preferences for location
of care: implications for regionalization. Med Care. 1999;37:204-209.
20. Marshall MN, Shekelle PG, Leatherman S, Brook RH. The public release of performance
data: what do we expect to gain? A review of the evidence JAMA. 2000;283:1866-1874.
211

21. Mennemeyer ST, Morrisey MA, Howard LZ. Death and reputation: how consumers acted
upon HCFA mortality information. Inquiry. 1997;34:117-128.
22. Khuri SF, Daley J, Henderson W, et al. The Department of Veterans Affairs' NSQIP: the
first national, validated, outcome-based, risk-adjusted, and peer-controlled program for the
measurement and enhancement of the quality of surgical care. National VA Surgical
Quality Improvement Program. Ann Surg. 1998;228:491-507.
23. Chassin MR, Hannan EL, DeBuono BA. Benefits and hazards of reporting medical
outcomes publicly. N Engl J Med. 1996;334:394-398.
24. Coronary Artery Bypass Surgery in New York State, 1992-1994 Albany, NY: New York
State Department of Health. 1996.
212

Chapter 19. Learning Curves for New Procedures – the Case of
Laparoscopic Cholecystectomy
Verna C. Gibbs, MD
Andrew D. Auerbach, MD, MPH
University of California, San Francisco School of Medicine
Background
Minimal access surgery began in the early 1980s with the introduction of laparoscopic
fallopian tube ligation. The first laparoscopic cholecystectomy was performed 7 years later, and
was rapidly embraced as the preferred method for cholecystectomy despite a lack of evidence to
support the safety of the new technique. 1-6 In response to several deaths and complications
associated with laparoscopic cholecystectomy, the New York State Department of Health issued
guidelines for the credentialing of surgeons who wished to perform the procedure. 7,8 At the same
time, a National Institutes of Health Consensus conference published recommendations
regarding indications for laparoscopic cholecystectomy. 1
Clinical trials comparing the laparoscopic procedure with other approaches eventually
revealed that the newer procedure to be less morbid than traditional open cholecystectomy, or
even mini-laparotomy. 9-16 Importantly, though, it also became clear that acquiring the skills to
perform this new procedure involved a substantial surgical “learning curve.” 17-22 This learning
curve no longer affects patients undergoing laparoscopic cholecystectomy, as training has
become a required part of all surgical residency programs, with graduating surgery residents
typically having performed more than 50 laparoscopic cholecystectomies. 23,24 However, the
growth of laparoscopic cholecystectomy has been followed by equally rapid development and
application of minimal access procedures in virtually every surgical specialty. This chapter
considers the patient safety issues that arise with the diffusion of a new procedurally-based
technology. It highlights the “learning curve” inherent in any new procedure, as competence
invariably grows with practice (see Chapter 18). Because it is so widely performed and has the
largest literature describing its record, this chapter focuses on lessons learned from the
introduction of laparoscopic cholecystectomy.
Practice Description
Although there are a wide variety of procedure-specific techniques in minimal access
surgery, all operations utilize videoscopic or digital imaging in conjunction with remote
manipulation. In general, the operator works from two-dimensional, magnified video images of
the operative field while manipulating long narrow instruments placed into the operating cavity
from outside the body. To become proficient in minimal access techniques, the surgeon must
develop skills in interpreting a three-dimensional environment as a two-dimensional image, and
learn how to do familiar tasks (eg, suture) with familiar instruments in an unfamiliar manner.
Importantly, the surgeon never actually touches the tissue being moved with his or her hands. 21,25
This loss of tactile input is the major factor in making minimal access techniques difficult to
learn.
213

Prevalence and Severity of the Target Safety Problem
Complications from minimal access procedures fall into 2 general categories: those
directly resulting from the laparoscopic intervention (ie, trocar injuries, dissection injuries,
insufflation-associated events) 26 and those associated with the operation itself (eg, bile duct
injury with cholecystectomy, gastric injury with fundoplication).* 17-22 For laparoscopic
cholecystectomy, a survey of 1750 surgery department chairpersons reported that bowel and
vascular injuries (laparoscopic-specific injury) occurred in 0.14% and 0.25% of cases
respectively, while the rate of bile duct injuries (procedure-specific) was 0.6%. 3 The bile duct
injury rate was lower at institutions that had performed more than 100 cases. Although this study
was large, as a survey it likely underestimates the true rate of complications. Other multiinstitutional
reports of laparoscopic cholecystectomy in the United States have noted a similar
bile duct injury rate and report mortality rates of 0.04% to 0.1%. 5,21,25,29 Although undoubtedly
underestimated, even these rates are higher than those associated with open cholecystectomy.
Opportunities for Impact
As set forth above, laparoscopic cholecystectomies is a standard part of surgical
residency training. 23,24 Nearly all general surgeons who did not train in the current era have
received postgraduate training in basic laparoscopic techniques. This training often takes the
form of a 2-3 day course involving hands-on experience with animal models and laboratoryskills
sessions, then observation or assisting with cases primarily performed by another surgeon,
finally, performing cases supervised by an experienced laparoscopist, and finally performing
cases independently. 6,24,30
Proficiency in traditional techniques does not automatically translate to minimal access
methods. We identified one study investigating whether the level of surgical training (attending
surgeon versus chief resident) affected the performance of cholecystectomy. 17 This study
suggested that, despite operating on more complex patients, the adverse event rate for surgical
chief residents learning laparoscopic cholecystectomy was similar to that of attending physicians
(who had far more operative experience).
Similarly, skills obtained with one minimal access procedure do not transfer to others.
The operation of laparoscopic fundoplication for gastroesophageal reflux disease entered into
surgical practice after laparoscopic cholecystectomy. Therefore, most surgeons who performed it
had basic laparoscopic skill sets. Studies investigating its implementation suggest a threshold of
25-30 cases before surgeons attain proficiency, 31-34 in contrast to a threshold of about 15-20
cases for laparoscopic cholecystectomy. 25,29,30,35 In recognition of this lack of transferability,
performing laparoscopic cholecystectomies does not qualify a surgeon to perform laparoscopic
colon resection without first undergoing additional training specifically in laparoscopic colon
techniques. 6,36 The need for specific new training has been seen with other procedures as well. 37-
41
* A third problem concerns the potential adverse effects at the population. While the procedure
itself may be less morbid, the rate at which this ‘safer’ procedure is performed may increase
substantially to the point that the complications experienced by the patient population as a whole
do not decrease. 27 This problem has been documented with laparoscopic cholecystectomy, 12,28
but processes for improved patient selection have not received sufficient testing to permit review
of a particular safety practice relating to procedure “appropriateness.”
214

Thus, determination of a minimum number of cases required to become proficient (as an
aid to credentialing or proctoring activities), or determining methods to shorten the time needed
to attain this level of skill will have a clear impact on patients undergoing minimal access
procedures. This impact is likely to become greater as the breadth and complexity of procedures
approached through minimal access techniques grows.
Study Designs
Using a structured MEDLINE search, we sought to identify papers that either reported a
threshold number of cases required to attain proficiency, or explicitly compared training
techniques for laparoscopic cholecystectomy in terms of patient outcomes. Our literature search
found a large number of reports describing retrospective analyses or reports from procedure
registries. These studies did not report training protocols, specific physician characteristics or
surgeon-specific volumes, and were therefore excluded from our review. In addition, we found
several review articles. 42-44 The first of these reviews focused primarily on statistical techniques
for documenting and analyzing learning curves, 42 and the other two 43,44 did not meet criteria for
systematic reviews. Nonetheless, the bibliographies of all 3 articles were scanned for relevant
studies.
Study Outcomes
All studies reported clinical outcomes (Level 1), most commonly bile duct injury. Other
complications (eg, insufflation or trocar injuries) were very unusual, and mortality was rare.
Evidence for Effectiveness of the Practice
Three studies specifically addressed the relationship between surgical experience in the
performance of laparoscopic cholecystectomy and a well-defined adverse outcome, bile duct
injury. The first reported 15 bile duct injuries in over 8000 laparoscopic cholecystectomies. 30
Multivariate analysis showed that the only significant factor in predicting this adverse event was
the experience of the surgeon (p=0.001). Rapidity of learning was not significantly related to a
surgeon's age, size of practice, or hospital setting. Ninety percent of injuries were predicted to
occur during a surgeon’s first 30 cases. Gigot et al 35 reported the incidence of bile duct injury
was 1.3% when the surgeon had performed fewer than 50 cases and 0.35% afterwards (p100 cases. 35
Similar results have been observed in other studies of bile duct injury with laparoscopic
cholecystectomy, 45 suggesting that something beyond the learning curve accounts for many
laparoscopic errors. Examination from error analysis suggests that these events occur because of
visual and perceptual difficulties involved in the operator/equipment interface. 46,47
We identified one prospective cohort study from Japan that compared the effectiveness of
2 training courses for laparoscopic cholecystectomy. 2 In this study, one group of 8 surgeons was
assigned to 10 supervised laparoscopic cholecystectomies as part of their initial training; a
second group of 8 surgeons had only 2 observed training sessions. Complications that occurred
over 21 months after completion of the 4 months of training were assessed by means of a
questionnaire sent to all participants. The surgeons who had trained with 10 supervised
procedures had 0.5% major complications (bleeding, bile duct injury or bile leakage) versus 2%
for the surgeons trained with only 2 procedures (p=0.03). The incidence of major complications
occurring during the initial 10 operations was higher among unsupervised surgeons (p=0.005).
Outside of laparoscopic cholecystectomy, we identified one study examining
complication rates of urologic laparoscopic surgeons after completing a 2-day training seminar. 48
215

Because the study relied on surveys of participants at 3 and 12 months after course completion, it
is likely biased toward underestimating the incidence of adverse events. Nevertheless the study
showed that surgeons who performed procedures without additional training were 3 times more
likely to have at least one complication compared with surgeons who sought additional training.
Additionally the presence of skilled associates and the development of a surgical team impact
favorably on reducing the number of complications. 48,49
Potential for Harm
It is unlikely that the specific introduction of training and requirements for a baseline of
surgical experience in minimal access surgery procedures will lead to adverse events.
Costs and Implementation
In the initial introduction of laparoscopic cholecystectomy most surgeons attended at
least one didactic and hands-on course in operative technique and instrument manipulation.
Some of these experiences included operative procedures in animals and utilization of static
training boxes and devices that may speed the acquisition of perceptual and motor skills. The
need for practice in the development of technical skills is vital, and laboratory courses, training
devices and simulators (Chapter 45) are being tested for their ability to improve operator
skills. 50-55 There is insufficient evidence to make a recommendation of the superiority of one
training method over any other and costs are presently undetermined.
After sufficient laboratory experience and practice, it is recommended that the procedure
be performed initially in only carefully selected patients under the supervision of an experienced
surgeon. 56-58 How practicing groups of providers obtain an experienced consultant is
variable. 59,60 Surgical training fellowships in minimal access surgery have been established and
graduates of these programs have the requisite expertise to supervise others. Unfortunately, such
fellowship graduates are in short supply. Telemedicine may play a role in the mentoring and
technical support for surgeons performing new techniques as it provides a means for remote
performance of a procedure with real-time expert supervision and guidance. 61,62
Comment
Minimal access surgery has become extraordinarily popular largely in response to market
forces. The single most important predictor of adverse events in minimal access procedures is
the experience of the provider with the specific operation. Surgeons must acquire the necessary
technical skills and expertise before performing new procedures on patients. Hospitals and
payers should help ensure that providers possess the requisite experience before allowing
procedures to be performed in their facilities or paying for them, since patients alone will
generally be unable to determine surgeon competency.
A number of governing bodies and surgical societies have published guidelines that
outline standards for training for postgraduate surgeons for skill acquisition in minimal access
surgery, 1,56-58,63-66 but these recommendations are based more on common sense and clinical
experience than rigorous evidence. It is not known how influential these guidelines are in the
granting of privileges. Continued research is needed to determine the threshold for safe
performance of this and other procedures, the most effective training methods to ensure
competence, and strategies to minimize patient harm while proceduralists gain the experience
they need to be competent and to train others.
216

References
1. Gallstones and laparoscopic cholecystectomy. NIH Consensus Development Panel on
Gallstones and Laparoscopic Cholecystectomy. Surg Endosc. 1993; 7: 271-279.
2. Yamashita Y, Kurohiji T, Kakegawa T. Evaluation of two training programs for
laparoscopic cholecystectomy: incidence of major complications. World J Surg.
1994;18:279-285.
3. Deziel DJ, Millikan KW, Economou SG, Doolas A, Ko ST, Airan MC. Complications of
laparoscopic cholecystectomy: a national survey of 4,292 hospitals and an analysis of
77,604 cases. Am J Surg. 1993;165:9-14.
4. Hodgson WJ, Byrne DW, Savino JA, Liberis G. Laparoscopic cholecystectomy. The early
experience of surgical attendings compared with that of residents trained by apprenticeship.
Surg Endosc. 1994; 8:1058-1062.
5. Wherry DC, Rob CG, Marohn MR, Rich NM. An external audit of laparoscopic
cholecystectomy performed in medical treatment facilities of the Department of Defense.
Ann Surg. 1994; 220:626-634.
6. Grundfest WS. Credentialing in an era of change. JAMA. 1993; 270: 2725.
7. NY State issues guidelines for lap cholecystectomy. New York State Health Department.
OR Manager. 1992;8:1,8-9.
8. Green FL. New York State Health Department ruling–a "wake-up call" for all. Surg
Endosc. 1992;6:271.
9. Shea JA, Healey MJ, Berlin JA, Clarke JR, Malet PF, Staroscik RN, et al. Mortality and
complications associated with laparoscopic cholecystectomy. A meta-analysis. Ann Surg.
1996;224:609-620.
10. Majeed AW, Troy G, Nicholl JP, Smythe A, Reed MW, Stoddard CJ, et al. Randomised,
prospective, single-blind comparison of laparoscopic versus small-incision
cholecystectomy. Lancet. 1996;347:989-994.
11. Kane RL, Lurie N, Borbas C, Morris N, Flood S, McLaughlin B, et al. The outcomes of
elective laparoscopic and open cholecystectomies. J Am Coll Surg. 1995;180:136-145.
12. Steiner CA, Bass EB, Talamini MA, Pitt HA, Steinberg EP. Surgical rates and operative
mortality for open and laparoscopic cholecystectomy in Maryland. N Engl J Med.
1994;330:403-408.
13. McMahon AJ, Russell IT, Baxter JN, Ross S, Anderson JR, Morran CG, et al.
Laparoscopic versus minilaparotomy cholecystectomy: a randomised trial. Lancet.
1994;343:135-138.
14. Barkun JS, Barkun AN, Meakins JL. Laparoscopic versus open cholecystectomy: the
Canadian experience. The McGill Gallstone Treatment Group. Am J Surg. 1993;165:455-
458.
15. Orlando R, Russell JC, Lynch J, Mattie A. Laparoscopic cholecystectomy. A statewide
experience. The Connecticut Laparoscopic Cholecystectomy Registry. Arch Surg.
1993;128:494-8.
16. Trondsen E, Reiertsen O, Andersen OK, Kjaersgaard P. Laparoscopic and open
cholecystectomy. A prospective, randomized study. Eur J Surg. 1993;159:217-221.
17. Lekawa M, Shapiro SJ, Gordon LA, Rothbart J, Hiatt JR. The laparoscopic learning curve.
Surg Laparosc Endosc. 1995;5:455-458.
217

18. Dashow L, Friedman I, Kempner R, Rudick J, McSherry C. Initial experience with
laparoscopic cholecystectomy at the Beth Israel Medical Center. Surg Gynecol Obstet.
1992;175:25-30.
19. Gilchrist BF, Vlessis AA, Kay GA, Swartz K, Dennis D. Open versus laparoscopic
cholecystectomy: an initial analysis. J Laparoendosc Surg. 1991;1:193-196.
20. Jones RM, Fletcher DR, MacLellan DG, Lowe AW, Hardy KJ. Laparoscopic
cholecystectomy: initial experience. Aust N Z J Surg. 1991;61:261-266.
21. Peters JH, Ellison EC, Innes JT, Liss JL, Nichols KE, Lomano JM, et al. Safety and
efficacy of laparoscopic cholecystectomy. A prospective analysis of 100 initial patients.
Ann Surg. 1991;213:3-12.
22. Sariego J, Spitzer L, Matsumoto T. The "learning curve" in the performance of
laparoscopic cholecystectomy. Int Surg. 1993;78:1-3.
23. Dent TL. Training and privileging for new procedures. Surg Clin North Am. 1996;76:615-
621.
24. Parsa CJ, Organ CH, Jr., Barkan H. Changing patterns of resident operative experience
from 1990 to 1997. Arch Surg. 2000;135:570-3.
25. Zucker KA, Bailey RW, Gadacz TR, Imbembo AL. Laparoscopic guided cholecystectomy.
Am J Surg. 1991;161:36-42.
26. Philips PA, Amaral JF. Abdominal access complications in laparoscopic surgery. J Am
Coll Surg. 2001;192:525-536.
27. Roos NP, Black CD, Roos LL, Tate RB, Carriere KC. A population-based approach to
monitoring adverse outcomes of medical care. Med Care. 1995;33:127-138.
28. Escarce JJ, Chen W, Schwartz JS. Falling cholecystectomy thresholds since the
introduction of laparoscopic cholecystectomy. JAMA. 1995;273:1581-1585.
29. A prospective analysis of 1518 laparoscopic cholecystectomies. The Southern Surgeons
Club. N Engl J Med. 1991;324:1073-1078.
30. Moore MJ, Bennett CL. The learning curve for laparoscopic cholecystectomy. The
Southern Surgeons Club. Am J Surg. 1995;170:55-59.
31. Champault GG, Barrat C, Rozon RC, Rizk N, Catheline JM. The effect of the learning
curve on the outcome of laparoscopic treatment for gastroesophageal reflux. Surg Laparosc
Endosc Percutan Tech. 1999;9:375-381.
32. Watson DI, Baigrie RJ, Jamieson GG. A learning curve for laparoscopic fundoplication.
Definable, avoidable, or a waste of time? Ann Surg. 1996;224:198-203.
33. Soot SJ, Eshraghi N, Farahmand M, Sheppard BC, Deveney CW. Transition from open to
laparoscopic fundoplication: the learning curve. Arch Surg. 1999;134:278-281.
34. Deschamps C, Allen MS, Trastek VF, Johnson JO, Pairolero PC. Early experience and
learning curve associated with laparoscopic Nissen fundoplication. J Thorac Cardiovasc
Surg. 1998;115:281-288.
35. Gigot J, Etienne J, Aerts R, Wibin E, Dallemagne B, Deweer F, et al. The dramatic reality
of biliary tract injury during laparoscopic cholecystectomy. An anonymous multicenter
Belgian survey of 65 patients. Surg Endosc. 1997;11:1171-1178.
36. Agachan F, Joo JS, Weiss EG, Wexner SD. Intraoperative laparoscopic complications. Are
we getting better? Dis Colon Rectum. 1996;39:S14-19.
37. Hunter JG. Advanced laparoscopic surgery. Am J Surg. 1997;173:14-18.
38. Janetschek G, Hobisch A, Holtl L, Bartsch G. Retroperitoneal lymphadenectomy for
clinical stage I nonseminomatous testicular tumor: laparoscopy versus open surgery and
impact of learning curve. J Urol. 1996;156:89-93.
218

39. Crawford DL, Phillips EH. Laparoscopic repair and groin hernia surgery. Surg Clin North
Am. 1998;78:1047-1062.
40. Meinke AK, Kossuth T. What is the learning curve for laparoscopic appendectomy? Surg
Endosc. 1994;8:371-375.
41. Perino A, Cucinella G, Venezia R, Castelli A, Cittadini E. Total laparoscopic hysterectomy
versus total abdominal hysterectomy: an assessment of the learning curve in a prospective
randomized study. Hum Reprod. 1999;14:2996-2999.
42. Ramsay CR, Grant AM, Wallace SA, Garthwaite PH, Monk AF, Russell IT. Assessment of
the learning curve in health technologies. A systematic review. Int J Technol Assess Health
Care. 2000;16:1095-1108.
43. Rogers DA, Elstein AS, Bordage G. Improving continuing medical education for surgical
techniques: applying the lessons learned in the first decade of minimal access surgery. Ann
Surg. 2001;233:159-166.
44. Hamdorf JM, Hall JC. Acquiring surgical skills. Br J Surg. 2000;87:28-37.
45. Steele RJ, Marshall K, Lang M, Doran J. Introduction of laparoscopic cholecystectomy in a
large teaching hospital: independent audit of the first 3 years. Br J Surg. 1995;82:968-971.
46. Olsen DO. Bile duct injuries during laparoscopic cholecystectomy: a decade of experience.
J Hepatobiliary Pancreat Surg. 2000;7:35-39.
47. Stewart L, Way LW. Bile duct injuries during laparoscopic cholecystectomy. Factors that
influence the results of treatment. Arch Surg. 1995;130:1123-1128.
48. See WA, Cooper CS, Fisher RJ. Predictors of laparoscopic complications after formal
training in laparoscopic surgery. JAMA. 1993;270:2689-2692.
49. Sexton JB, Thomas EJ, Helmreich RL. Error, stress, and teamwork in medicine and
aviation: cross sectional surveys. BMJ. 2000;320:745-749.
50. Noar MD. The next generation of endoscopy simulation: minimally invasive surgical skills
simulation. Endoscopy. 1995;27:81-85.
51. Derossis AM, Fried GM, Abrahamowicz M, Sigman HH, Barkun JS, Meakins JL.
Development of a model for training and evaluation of laparoscopic skills. Am J Surg.
1998;175:482-487.
52. Keyser EJ, Derossis AM, Antoniuk M, Sigman HH, Fried GM. A simplified simulator for
the training and evaluation of laparoscopic skills. Surg Endosc. 2000;14:149-153.
53. Scott DJ, Bergen PC, Rege RV, Laycock R, Tesfay ST, Valentine RJ, et al. Laparoscopic
training on bench models: better and more cost effective than operating room experience? J
Am Coll Surg. 2000;191:272-283.
54. Rosser JC, Herman B, Risucci DA, Murayama M, Rosser LE, Merrell RC. Effectiveness of
a CD-ROM multimedia tutorial in transferring cognitive knowledge essential for
laparoscopic skill training. Am J Surg. 2000;179:320-324.
55. Gallagher AG, McClure N, McGuigan J, Crothers I, Browning J. Virtual reality training in
laparoscopic surgery: a preliminary assessment of minimally invasive surgical trainer
virtual reality (MIST VR). Endoscopy. 1999;31:310-313.
56. Guidelines for the clinical application of laparoscopic biliary tract surgery. Society of
American Gastrointestinal Endoscopic Surgeons. Surg Endosc. 2000;14:771-772.
57. Guidelines for granting of privileges for laparoscopic and/or thoracoscopic general
surgery. Society of American Gastrointestinal Endoscopic Surgeons (SAGES). Surg
Endosc. 1998;12:379-380.
58. Guidelines for surgical treatment of gastroesophageal reflux disease (GERD). Society of
American Gastrointestinal Endoscopic Surgeons (SAGES). Surg Endosc. 1998;12:186-188.
219

59. Fowler DL, Hogle N. The impact of a full-time director of minimally invasive surgery:
clinical practice, education, and research. Surg Endosc. 2000;14:444-447.
60. Sequeira R, Weinbaum F, Satterfield J, Chassin J, Mock L. Credentialing physicians for
new technology: the physician's learning curve must not harm the patient. Am Surg.
1994;60:821-823.
61. Deaton DH, Balch D, Kesler C, Bogey WM, Powell CS. Telemedicine and endovascular
aortic grafting. Am J Surg. 1999;177:75-77.
62. Moore RG, Adams JB, Partin AW, Docimo SG, Kavoussi LR. Telementoring of
laparoscopic procedures: initial clinical experience. Surg Endosc. 1996;10:107-110.
63. Guidelines for submission of continuing medical education seeking SAGES endorsement
for courses in laparoscopic surgery. Society of American Gastrointestinal Endoscopic
Surgeons (SAGES). Surg Endosc. 1993;7:372-373.
64. Guidelines for diagnostic laparoscopy.SAGES guidelines. Society of American
Gastrointestinal Endoscopic Surgeons. Surg Endosc. 1999;13:202-203.
65. Guidelines for laparoscopic surgery during pregnancy. Surg Endosc. 1998;12:189-190.
66. Guidelines for office endoscopic services. Society of American Gastrointestinal
Endoscopic Surgeons (SAGES). Surg Endosc. 1998;12:191-192.
220

Chapter 20. Prevention of Surgical Site Infections
Andrew D. Auerbach, MD, MPH
University of California, San Francisco School of Medicine
Subchapter 20.1. Prophylactic Antibiotics
Background
Surgical site infections (SSI) include superficial incisional infections, infections of the
deep incision space and organ space infections. 1,2 A large body of evidence supports the premise
that SSIs can be prevented through administration of appropriate prophylactic antibiotics. Two
national organizations, the Federal Centers for Disease Control and Prevention (CDC) and the
American Society for Health System Pharmacists (ASHP), have recently synthesized this vast
literature to produce comprehensive guidelines regarding the administration of prophylactic
antibiotics across a broad range of procedures. 3,4 Because of the breadth of this literature, we
limited the focus of this review of strategies to prevent SSIs to adult surgery and procedures that
typically occur in the operating room (as opposed to procedures such as endoscopy,
interventional cardiology, or radiology procedures).
Practice Description
Antimicrobial prophylaxis refers to a brief course of an antimicrobial agent administered
just before an operation begins in order to reduce intraoperative microbial contamination to a
level that will not overwhelm host defenses and result in infection. 4 To maximize the benefits of
antimicrobial prophylactic, the agent used should be safe, inexpensive, and bactericidal with an
in vitro spectrum that covers the most probable contaminants for the operation. 4 Administration,
usually by intravenous infusion, should be timed so that a bactericidal concentration is present in
serum and tissues by the time the skin is incised. 5 This practice is now standard of care and
recommended by professional societies. 6 Therapeutic levels in serum and tissues should be
maintained until, at most, a few hours after the incision is closed in the operating room. 4
Prevalence and Severity of the Target Safety Problem
Surgical site infections are a common complication of care, occurring in 2-5% of patients
after clean extra-abdominal surgeries (eg, thoracic and orthopedic surgery) and in up to 20% of
patients undergoing intra-abdominal procedures. 7-12 Studies following patients into the postdischarge
period have reported even higher rates of postoperative infection. 13-16 These
complications increase morbidity for patients and consume substantial additional resources. 17-21
Opportunities for Impact
Approximately 80-90% of surgical patients receive some kind of antibiotic prophylaxis,
though recent studies have shown that choice of regimen, timing of administration or duration of
prophylaxis is inappropriate in approximately 25-50% of cases. 22-27
221

Study Designs and Outcomes
As previously noted, the literature on prophylactic antibiotics is extensive. Therefore, the
review was limited to evidence from Level 1A study designs. We identified 9 relevant studies
examining the use of prophylactic antibiotics to prevent surgical site infections: 7 meta-analyses
and 2 systematic reviews. 28-36 (Tables 20.1.1 and 20.1.2) These reviews were of high quality and
limited their source material to randomized controlled trials. Although additional randomized
trials have been published since these reviews were performed, updating the results of each
review was beyond the scope of this project. All studies examined measured rates of site
infection directly (Level 1), using previously published definitions to allow comparability. In
addition, the rates of sepsis, length of stay, and physiologic measures were reported. One metaanalysis
31 and one systematic review 33 combined rates of several relevant infectious outcomes.
Evidence for Effectiveness of the Practice
All studies showed a marked reduction in the odds or relative risk of SSI when antibiotic
prophylaxis was employed. None of the meta-analyses reviewed explicitly examined the timing
of prophylaxis, although many studies pooled data from investigations of antibiotic regimens
administered in the immediate preoperative period, (ie, within minutes to an hour of initial
incision). Two meta-analyses in our review 29,31 suggested a trend towards lower rates of
infection with use of broader-spectrum antibiotic prophylaxis, such as third generation
cephalosporins. When compared with single dose prophylaxis, multiple dose prophylaxis
generally did not result in significant additional benefit. 29,30,35 In fact, Tanos et al found the odds
of SSI were significantly less with single dose prophylaxis. 31 Gillespie et al reported a greater
relative risk of infection with single dose prophylaxis with a short-acting antibiotic when
compared with multiple dose prophylaxis. 36 However, the risk of infection with single dose
prophylaxis using long-acting antibiotics did not differ significantly from that seen with
multiple-dose regimens.
Potential for Harm
None of the meta-analyses analyzed reported rates of adverse events (such as allergic
reactions or nosocomial infections) associated with antibiotic prophylaxis of any type or
duration. Both of the systematic reviews 33,36 noted a trend towards more frequent adverse events
with the use of antibiotic prophylaxis. Authors of both systematic reviews observed that these
events were reported rarely and that variation in the definition of “adverse events” across studies
made pooling results difficult.
Infection with Clostridium difficile affects a large number of hospitalized patients and has
significant clinical and economic implications. As many as 16% of C. difficile colitis cases in
surgical patients can be attributed to prophylaxis alone, 37 with higher risk for this complication
among patients receiving broad-spectrum antibiotics or prolonged courses of therapy. Shortening
the duration of antibiotic administration may reduce potential risks of prophylaxis (see Chapter
14). Emergence of other types of resistant pathogens is an additional theoretical concern of
inappropriate antibiotic prophylaxis; our literature search found no data describing effect of
antibiotic prophylaxis on population-level incidence of these pathogens.
222

Costs and Implementation
A number of studies have evaluated strategies for improving compliance with
recommended practices for perioperative antibiotic prophylaxis. These include chart audit with
feedback, 38 computerized decision support, 23, 39-42 dissemination of guidelines, 43 total quality
management (TQM) and continuous quality improvement (CQI) techniques, 44-47 provider
education programs, 48,49 and comprehensive efforts by an infection control team. 50 Another
promising and easily implemented method is to delegate the administration of prophylactic
22, 25, 48
antibiotics to the anesthesia team or the holding room nursing staff.
Costs for systems to increase appropriate use of antibiotics will likely be offset by
savings due to prevented infections. However formal analyses of the cost-effectiveness of
specific programs to improve prophylaxis have not been reported.
Comment
For many surgical procedures there is clear evidence supporting the use of antibiotic
prophylaxis, administered in a timely manner, to prevent surgical site infections. The reviews
suggest that broader spectrum antibiotics may be superior to limited-spectrum antibiotics for
intra-abdominal or gynecologic surgeries. In addition, single-dose antibiotic prophylaxis appears
to be at least as effective as multiple-dose regimens for a broad range of surgical procedures and
may pose less risk to patients in terms of adverse events (eg, C. difficile colitis) and less risk to
the population in terms of microbial resistance.
Future research will continue to address what prophylactic regimens are most effective
for various surgical procedures. Investigation should also focus on methods to improve
compliance. The optimal strategies for implementation will likely vary from institution to
institution.
223

Table 20.1.1. Meta-analyses examining antibiotic prophylaxis*
Study Trials
Included
Kreter,
1992 35
McDonal
d, 1998 30
Meijer,
1990 29
Mittendor
f, 1993 28
Sharma,
2000 34
Tanos,
1994 31
Surgical Procedures,
Antibiotics
28 Cardiothoracic surgery;
cephalosporins
28 Multiple types of
surgery;
multiple antibiotics
42 Biliary surgery;
cephalosporins
25 Abdominal
hysterectomy;
multiple antibiotics
6 Percutaneous
gastrostomy;
multiple antibiotics
17 Abdominal
hysterectomy;
cephalosporins
Results: Odds Ratio or Relative Risk of
Infection (95% CI)
ƒ Cefazolin vs. placebo: OR 0.2 (0.10-0.48).
ƒ Cefazolin vs. cefuroxime or cefamandole:
OR 1.6 (1.03-2.45)
• Single dose vs. multiple dose regimen: no
significant difference
ƒ Single dose vs. multiple dose antibiotics (all
studies): OR 1.06 (0.89-1.25)
ƒ Duration of multiple dose regimen 24
hours: OR 1.08 (0.86-1.36)
ƒ Antibiotic vs. placebo: OR 0.30 (0.23-0.38)
ƒ Cephalosporin I vs. cephalosporin II or III:
OR 1.18 (0.69-2)†
• Single dose vs. multiple dose regimen: OR
0.80 (0.4-1.6)
ƒ Antibiotic vs. placebo (all studies): OR 0.35
(0.27-0.5); p

Wilson,
1992 51
21 Multiple types of
surgery;
multiple antibiotics
225
• Single dose vs. multiple dose regimen: OR
0.37 (0.3-0.5)
ƒ Amoxicillin-clavulanic acid vs. other
antibiotics (all studies): OR 0.84 (0.68-
1.04)
ƒ Trend favoring amoxicillin-clavulanic acid
for biliary and gynecologic surgery
* CI indicates confidence interval; NNT, number needed to treat; OR, odds ratio, and RR,
relative risk.
† Roman numerals I, II, III indicate generation of cephalosporin antibiotics.
‡ P values were reported in article; OR were approximated based on figures.

Table 20.1.2. Systematic reviews of antibiotic prophylaxis*
Study Trials
Included
Gillespie,
2000 36
Smaill,
2000 33
Surgical
Procedures;
Antibiotics
Results: Relative Risk of Infection (95% CI)
48 Long bone Single dose antibiotic vs. placebo
fractures;
Deep wound infection: RR 0.40 (0.24-0.67)
multiple
antibiotics
Superficial wound infection: RR 0.69 (0.50-0.95)
Urinary tract infection: RR 0.63 (0.53-0.76)
Pneumonia: RR 0.46 (0.33-0.65)
Multiple dose antibiotic vs. placebo:
Deep wound infection: RR 0.36 (0.21-0.65)
Superficial wound infection: RR 0.48 (0.28-0.81)
Urinary tract infection: RR 0.66 (0.4-1.0)
Pneumonia: RR 0.81 (0.41-1.63)
Adverse events: RR 1.83 (0.96-3.50)
Single dose short-acting antibiotic vs. multiple doses same
agent up to 24 hours after surgery
Deep wound infection: RR 7.98 (1.01-62.0)
Superficial wound infection: RR 4.82 (1.08-21.6)
Urinary tract infection: RR 1.81 (1.01-3.23)
Single dose long-acting antibiotic vs. any multiple dose
regimen lasting more than 24 hours
Deep wound infection: RR 1.10 (0.22-5.34)
Superficial wound infection: RR 0.57 (0.17-1.93)
Multiple doses administered over 24 hours or less vs.
longer therapy
Deep wound infection: RR 1.1 (0.22-5.34)
Superficial wound infection: RR 0.57 (0.17-1.93)
Oral vs. parenteral prophylaxis
Insufficient data (single underpowered study)
66 Cesarean section; Impact of antibiotic prophylaxis on …
multiple -Combined outcomes of fever, wound infection, sepsis and
antibiotics endometritis:
Elective Cesarean section: RR 0.25 (0.11-0.55)
Emergent Cesarean section: RR 0.39 (0.33-0.46)
Unspecified/nonelective: RR 0.37 (0.32-0.42)
226

All Cesarean section: RR 0.37 (0.33-0.42)
-Maternal side effects: RR 1.96 (0.86-4.49)
-Length of stay: 0.34 fewer days in hospital (0.17-0.52)
* CI indicates confidence interval; RR, relative risk.
227

References
1. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for the Prevention
of Surgical Site Infection. Hospital Infection Program, Centers for Disease Control and
Prevention. Available at: http://www.cdc.gov/ncidod/hip/. Accessed April 29, 2001.
2. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for prevention of
surgical site infection, 1999. Hospital Infection Control Practices Advisory Committee.
Infect Control Hosp Epidemiol. 1999;20:250-278.
3. ASHP Therapeutic Guidelines on Antimicrobial Prophylaxis in Surgery. American Society
of Health-System Pharmacists. Am J Health Syst Pharm. 1999;56:1839-1888.
4. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for Prevention of
Surgical Site Infection, 1999. Centers for Disease Control and Prevention (CDC) Hospital
Infection Control Practices Advisory Committee. Am J Infect Control. 1999;27:97-132.
5. Classen DC, Evans RS, Pestotnik SL, Horn SD, Menlove RL, Burke JP. The timing of
prophylactic administration of antibiotics and the risk of surgical-wound infection. N Engl
J Med. 1992;326:281-286.
6. Dellinger EP, Gross PA, Barrett TL, et al. Quality standard for antimicrobial prophylaxis in
surgical procedures. The Infectious Diseases Society of America. Infect Control Hosp
Epidemiol. 1994;15:182-188.
7. Delgado-Rodriguez M, Sillero-Arenas M, Medina-Cuadros M, Martinez-Gallego G.
Nosocomial infections in surgical patients: comparison of two measures of intrinsic patient
risk. Infect Control Hosp Epidemiol. 1997;18:19-23.
8. Horan TC, Culver DH, Gaynes RP, Jarvis WR, Edwards JR, Reid CR. Nosocomial
infections in surgical patients in the United States, January 1986-June 1992. National
Nosocomial Infections Surveillance (NNIS) System. Infect Control Hosp Epidemiol.
1993;14:73-80.
9. Horan TC, Gaynes RP, Martone WJ, Jarvis WR, Emori TG. CDC definitions of
nosocomial surgical site infections, 1992: a modification of CDC definitions of surgical
wound infections. Infect Control Hosp Epidemiol. 1992;13:606-608.
10. Horan TC, Emori TG. Definitions of key terms used in the NNIS System. Am J Infect
Control. 1997;25:112-116.
11. Wallace WC, Cinat M, Gornick WB, Lekawa ME, Wilson SE. Nosocomial infections in
the surgical intensive care unit: a difference between trauma and surgical patients. Am
Surg. 1999;65:987-990.
12. Scheel O, Stormark M. National prevalence survey on hospital infections in Norway. J
Hosp Infect. 1999;41:331-335.
13. Sands K, Vineyard G, Platt R. Surgical site infections occurring after hospital discharge. J
Infect Dis. 1996;173:963-970.
14. Waddell TK, Rotstein OD. Antimicrobial prophylaxis in surgery. Committee on
Antimicrobial Agents, Canadian Infectious Disease Society. CMAJ. 1994;151:925-931.
15. Medina-Cuadros M, Sillero-Arenas M, Martinez-Gallego G, Delgado-Rodriguez M.
Surgical wound infections diagnosed after discharge from hospital: epidemiologic
differences with in-hospital infections. Am J Infect Control. 1996;24:421-428.
16. Lynch W, Davey PG, Malek M, Byrne DJ, Napier A. Cost-effectiveness analysis of the use
of chlorhexidine detergent in preoperative whole-body disinfection in wound infection
prophylaxis. J Hosp Infect. 1992;21:179-191.
228

17. Vegas AA, Jodra VM, Garcia ML. Nosocomial infection in surgery wards: a controlled
study of increased duration of hospital stays and direct cost of hospitalization. Eur J
Epidemiol. 1993;9:504-510.
18. Kirkland KB, Briggs JP, Trivette SL, Wilkinson WE, Sexton DJ. The impact of surgicalsite
infections in the 1990s: attributable mortality, excess length of hospitalization, and
extra costs. Infect Control Hosp Epidemiol. 1999;20:725-730.
19. Collins TC, Daley J, Henderson WH, Khuri SF. Risk factors for prolonged length of stay
after major elective surgery. Ann Surg. 1999;230:251-259.
20. Asensio A, Torres J. Quantifying excess length of postoperative stay attributable to
infections: a comparison of methods. J Clin Epidemiol. 1999;52:1249-1256.
21. Wong ES. The price of a surgical-site infection: more than just excess length of stay. Infect
Control Hosp Epidemiol. 1999;20:722-724.
22. Silver A, Eichorn A, Kral J, et al. Timeliness and use of antibiotic prophylaxis in selected
inpatient surgical procedures. The Antibiotic Prophylaxis Study Group. Am J Surg.
1996;171:548-552.
23. Larsen RA, Evans RS, Burke JP, Pestotnik SL, Gardner RM, Classen DC. Improved
perioperative antibiotic use and reduced surgical wound infections through use of computer
decision analysis. Infect Control Hosp Epidemiol. 1989;10:316-320.
24. Finkelstein R, Reinhertz G, Embom A. Surveillance of the use of antibiotic prophylaxis in
surgery. Isr J Med Sci. 1996;32:1093-1097.
25. Matuschka PR, Cheadle WG, Burke JD, Garrison RN. A new standard of care:
administration of preoperative antibiotics in the operating room. Am Surg. 1997;63:500-
503.
26. Gorecki P, Schein M, Rucinski JC, Wise L. Antibiotic administration in patients
undergoing common surgical procedures in a community teaching hospital: the chaos
continues. World J Surg. 1999;23:429-432.
27. Zoutman D, Chau L, Watterson J, Mackenzie T, Djurfeldt M. A Canadian survey of
prophylactic antibiotic use among hip-fracture patients. Infect Control Hosp Epidemiol.
1999;20:752-755.
28. Mittendorf R, Aronson MP, Berry RE, et al. Avoiding serious infections associated with
abdominal hysterectomy: a meta-analysis of antibiotic prophylaxis. Am J Obstet Gynecol.
1993;169:1119-1124.
29. Meijer WS, Schmitz PI, Jeekel J. Meta-analysis of randomized, controlled clinical trials of
antibiotic prophylaxis in biliary tract surgery. Br J Surg. 1990;77:283-290.
30. McDonald M, Grabsch E, Marshall C, Forbes A. Single- versus multiple-dose
antimicrobial prophylaxis for major surgery: a systematic review. Aust N Z J Surg.
1998;68:388-396.
31. Tanos V, Rojansky N. Prophylactic antibiotics in abdominal hysterectomy. J Am Coll Surg.
1994;179:593-600.
32. Song F, Glenny AM. Antimicrobial prophylaxis in colorectal surgery: a systematic review
of randomized controlled trials. Br J Surg. 1998;85:1232-1241. Published erratum appears
in Br J Surg 1999;1286:1280.
33. Smaill F, Hofmeyr GJ. Antibiotic prophylaxis for cesarean section. In: The Cochrane
Library, Issue 2, 2000. Oxford: Update Software.
34. Sharma VK, Howden CW. Meta-analysis of randomized, controlled trials of antibiotic
prophylaxis before percutaneous endoscopic gastrostomy. Am J Gastroenterol.
2000;95:3133-3136.
229

35. Kreter B, Woods M. Antibiotic prophylaxis for cardiothoracic operations. Meta-analysis of
thirty years of clinical trials. J Thorac Cardiovasc Surg. 1992;104:590-599.
36. Gillespie WJ, Walenkamp G. Antibiotic prophylaxis for surgery for proximal femoral and
other closed long bone fractures. In: The Cochrane Library, Issue 2, 2000. Oxford: Update
Software; 2000.
37. Crabtree TD, Pelletier SJ, Gleason TG, Pruett TL, Sawyer RG. Clinical characteristics and
antibiotic utilization in surgical patients with Clostridium difficile-associated diarrhea. Am
Surg. 1999;65:507-511.
38. Zhanel GG, Gin AS, Przybylo A, Louie TJ, Otten NH. Effect of interventions on
prescribing of antimicrobials for prophylaxis in obstetric and gynecologic surgery. Am J
Hosp Pharm. 1989;46:2493-2496.
39. Pestotnik SL, Evans RS, Burke JP, Gardner RM, Classen DC. Therapeutic antibiotic
monitoring: surveillance using a computerized expert system. Am J Med. 1990;88:43-48.
40. Pestotnik SL, Classen DC, Evans RS, Burke JP. Implementing antibiotic practice
guidelines through computer-assisted decision support: clinical and financial outcomes.
Ann Intern Med. 1996;124:884-890.
41. Evans RS, Pestotnik SL, Classen DC, et al. A computer-assisted management program for
antibiotics and other antiinfective agents. N Engl J Med. 1998;338:232-238.
42. Evans RS, Pestotnik SL, Burke JP, Gardner RM, Larsen RA, Classen DC. Reducing the
duration of prophylactic antibiotic use through computer monitoring of surgical patients.
DICP. 1990;24:351-354.
43. Dobrzanski S, Lawley DI, McDermott I, Selby M, Ausobsky JR. The impact of guidelines
on peri-operative antibiotic administration. J Clin Pharm Ther. 1991;16:19-24.
44. Koska MT. Using CQI methods to lower postsurgical wound infection rate. Hospitals.
1992;66:62,64.
45. Kroll DA, Brummitt CF, Berry BB. A users group approach to quality improvement across
an integrated healthcare delivery system. J Healthc Qual. 2000;22:39-43.
46. Welch L, Teague AC, Knight BA, Kenney A, Hernandez JE. A quality management
approach to optimizing delivery and administration of preoperative antibiotics. Clin
Perform Qual Health Care. 1998;6:168-171.
47. Woster PS, Ryan ML, Ginsberg-Evans L, Olson J. Use of total quality management
techniques to improve compliance with a medication use indicator. Top Hosp Pharm
Manage. 1995;14:68-77.
48. Gyssens IC, Knape JT, Van Hal G, ver der Meer JW. The anaesthetist as determinant factor
of quality of surgical antimicrobial prophylaxis. A survey in a university hospital. Pharm
World Sci. 1997;19:89-92.
49. Everitt DE, Soumerai SB, Avorn J, Klapholz H, Wessels M. Changing surgical
antimicrobial prophylaxis practices through education targeted at senior department
leaders. Infect Control Hosp Epidemiol. 1990;11:578-583.
50. McConkey SJ, L'Ecuyer PB, Murphy DM, Leet TL, Sundt TM, Fraser VJ. Results of a
comprehensive infection control program for reducing surgical-site infections in coronary
artery bypass surgery. Infect Control Hosp Epidemiol. 1999;20:533-538.
51. Wilson AP, Shrimpton S, Jaderberg M. A meta-analysis of the use of amoxycillinclavulanic
acid in surgical prophylaxis. J Hosp Infect. 1992;22:9-21.
230

Subchapter 20.2. Perioperative Normothermia
Background
The body temperature of patients may fall by 1 to 1.5°C during the first hour of general
anesthesia. 1 Regional anesthesia also typically causes core hypothermia. 2 Intraoperative
hypothermia impairs immune function (especially oxidative killing by neutrophils) and results in
dermal vasoconstriction and reduced blood flow to surgical sites, which further increases the risk
of surgical site infection by lowering tissue oxygen tension. 3 Hypothermia also results in reduced
platelet function, shivering associated with patient discomfort and activation of the sympathetic
nervous system, and adverse cardiac events. 2
Practice Description
Normal core temperature can be maintained during surgery through use of active
measures including warmed intravenous fluids and inspired gases, as well as forced air warming.
The latter involves an air blanket placed over the patient that circulates air warmed to 40°C.
Water blankets may also be used, but are not as effective in maintaining body temperature. 4
Patient temperature is monitored using conventional thermometer probes, with active measures
adjusted to maintain core temperature near 36.5°C. Any method or combination of methods that
maintains the target core temperature appears to have the same effect. 2
Prevalence and Severity of the Target Safety Problem
See Subchapter 20.1.
Opportunities for impact
Attention to patient temperature is standard of care in intraoperative anesthesia
management. * However, there are no data on the extent to which active warming measures are
currently used perioperatively.
* The American Society of Anesthesiologists’ Standards for Basic Anesthesia Monitoring notes
“Every patient receiving anesthesia shall have temperature monitored when clinically significant
changes in body temperature are intended, anticipated or suspected.” 5
231

Study Designs and Outcomes
We identified one randomized controlled trial 3 and one retrospective cohort study 6
evaluating the effect of active warming interventions on the rate of wound infection (Level 1
outcome). (Table 20.2.1). Wound infection was either defined as “suppuration requiring removal
of sutures” 3 or as in previously published definitions. 7
Evidence for Effectiveness of the Practice
Kurz et al performed a randomized controlled trial of active warming in the
intraoperative care of patients undergoing elective colectomy. All patients received aggressive
perioperative hydration and intravenous opioids for pain relief, in an effort to maximize wound
perfusion. Patients in the normothermia arm experienced a 68% reduction in the rate of wound
infection, lower wound infection scores (as defined by the elements of the acronym ASEPSIS:
Additional treatment, Serous discharge, Erythema, Purulent exudate, Separation of deep tissues,
Isolation of bacteria, and duration of inpatient Stay), and shorter length of hospitalization. 3
While the relatively high infection rate (19% of control group in this university-based population
with a substantial degree of underlying disease) and suboptimal antibiotic prophylaxis
(antibiotics continued for about 4 days postoperatively; see Subchapter 20.1) do not invalidate
the study results, they do limit their generalizability.
In a retrospective cohort study based on chart reviews of 150 patients undergoing elective
colectomy, Barone et al noted no independent association between intraoperative hypothermia
(defined as temperature less than

Comment
Given the evidence of effectiveness, the low potential for harm, and the simplicity of the
intervention (including the ready availability of the equipment), maintenance of perioperative
normothermia seems a promising practice to improve patient safety. The methodologically
stronger of the 2 studies reviewed showed clear benefits. However, some of its benefits may not
be generalizable to patient populations undergoing other procedures. For example, intraoperative
hypothermia may have little impact on wound infections in patients undergoing cesarean
section. 14 Thus, additional study of the practice is needed in other settings. Furthermore, for
some procedures hypothermia is likely to protect patients. Core temperature is often intentionally
reduced to protect the myocardium and central nervous system during certain cardiac and
neurosurgical procedures. 2,12,15 In such cases the potential benefits of normothermia may not
outweigh the associated risks.
Table 20.2.1. Summary of studies reporting effectiveness of perioperative normothermia*
Study Study Population; Intervention Study Design,
Outcomes
Kurz,
1996 3
Barone,
1999 6
200 patients (104 normothermia, 96
hypothermia) undergoing, elective
colectomy in multicenter study;
warmed gases, fluids and forced arm
during operation vs. usual care
150 patients (101 normothermia, 49
hypothermia) undergoing elective
colectomy at a single community
hospital; no formal intervention
(retrospective chart review,
warming devices were used in 90%
of patients)
Level 1,
Level 1
Level 3,
Level 1
233
Results
Wound infection rate: 6% vs. 19%
(p=0.009)
ASEPSIS score: 7 vs. 13 (p=0.002)
Days to sutures out: 9.9 vs. 10.9
(p=0.002)
Taking nutrition orally: 5.6 vs. 6.5 days
(p=0.006)
Length of stay: 12 vs. 15 days (p=0.001)
Wound infection rate: 12% in both
groups
Multivariate models: no significant
association between hypothermia
and wound infection or length of
stay
* ASEPSIS indicates Additional treatment, Serous discharge, Erythema, Purulent exudate, Separation of
deep tissues, Isolation of bacteria, and duration of inpatient Stay. 7
References
1. Matsukawa T, Sessler DI, Sessler AM, et al. Heat flow and distribution during induction of
general anesthesia. Anesthesiology. 1995;82:662-673.
2. Sessler DI. Mild perioperative hypothermia. N Engl J Med. 1997;336:1730-1737.
3. Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce the incidence of
surgical-wound infection and shorten hospitalization. Study of Wound Infection and
Temperature Group. N Engl J Med. 1996;334:1209-1215.

4. Kurz A, Kurz M, Poeschl G, Faryniak B, Redl G, Hackl W. Forced-air warming maintains
intraoperative normothermia better than circulating-water mattresses. Anesth Analg.
1993;77:89-95.
5. American Society of Anesthesiologists. Standards of the American Society of
Anesthesiologists. Available at: http://www.asahq.org/Standards/02.html. Accessed June 6,
2001.
6. Barone JE, Tucker JB, Cecere J, et al. Hypothermia does not result in more complications
after colon surgery. Am Surg. 1999;65:356-359.
7. Wilson AP, Treasure T, Sturridge MF, Gruneberg RN. A scoring method (ASEPSIS) for
postoperative wound infections for use in clinical trials of antibiotic prophylaxis. Lancet.
1986;1:311-313.
8. Sessler DI, Kurz A, Lenhardt R. Re: Hypothermia reduces resistance to surgical wound
infections. Am Surg. 1999;65:1193-1196.
9. Frank SM, Fleisher LA, Breslow MJ, et al. Perioperative maintenance of normothermia
reduces the incidence of morbid cardiac events. A randomized clinical trial. JAMA.
1997;277:1127-1134.
10. Schmied H, Kurz A, Sessler DI, Kozek S, Reiter A. Mild hypothermia increases blood loss
and transfusion requirements during total hip arthroplasty. Lancet. 1996;347:289-292.
11. Winkler M, Akca O, Birkenberg B, et al. Aggressive warming reduces blood loss during
hip arthroplasty. Anesth Analg. 2000;91:978-984.
12. Leslie K, Sessler DI. The implications of hypothermia for early tracheal extubation
following cardiac surgery. J Cardiothorac Vasc Anesth. 1998;12:30-34; discussion 41-34.
13. Sigg DC, Houlton AJ, Iaizzo PA. The potential for increased risk of infection due to the
reuse of convective air-warming/cooling coverlets. Acta Anaesthesiol Scand. 1999;43:173-
176.
14. Munn MB, Rouse DJ, Owen J. Intraoperative hypothermia and post-cesarean wound
infection. Obstet Gynecol. 1998;91:582-584.
15. Winfree CH, Baker KZ, Connollly ES. Perioperative normothermia and surgical-wound
infection. N Engl J Med. 1996;335:749.
Subchapter 20.3. Supplemental Perioperative Oxygen
Background
Low oxygen content in devitalized tissues predisposes them to bacterial colonization,
which is thought to be a key pathophysiologic step in the initiation of surgical site infections. 1
Administration of high concentrations of oxygen increases wound oxygen tension, allowing for
more effective neutrophil function and the potential for reduced infection rates. 2
Practice Description
The practice of perioperative oxygen supplementation involves administration of 80%
oxygen and 20% nitrogen by endotracheal tube intraoperatively and by sealed mask and
manifold system or conventional non-rebreather mask for the first two hours of recovery.
Oxygen is increased to 100% immediately before extubation, with the concentration returned to
80% as soon as deemed safe by the anesthesiologist. 3
Prevalence and Severity of the Target Safety Problem
See Subchapter 20.1.
234

Opportunities for Impact
Administration of oxygen is a routine part of perioperative care. However the frequency
with which high oxygen concentrations (as described above) are administered is not known.
Study Designs and Outcomes
We identified one randomized controlled trial evaluating the effect of high concentration
oxygen supplementation on surgical site infections (Table 20.3.1). 3 The primary outcome was
incidence of wound infection within 15 days after surgery (Level 1). Wounds were considered
infected when bacteria were cultured from pus expressed from the incision or aspirated from a
loculated collection within the wound. 3
Evidence for Effectiveness of the Practice
The clinical characteristics of the intervention and control groups were similar at
baseline, including risk of infection as assessed by a modified Study on the Efficacy of
Nosocomial Infection Control (SENIC) score (p=0.8) and National Nosocomial Infection
Surveillance System (NNISS) score (p=0.86). The incidence of wound infection was
significantly less in the intervention group (13/250, 5%) than in the control group (28/250, 11%,
p=0.014). The results remain statistically significant when the study definition of “infection” is
broadened to include wounds with pus but no bacterial growth on culture (7% vs. 14%,
p=0.012). Perioperative administration of high levels of oxygen was associated with a 54%
relative risk reduction (95% CI: 12%-75%) of wound infection within 15 days of surgery.
ASEPSIS (Additional treatment, Serous discharge, Erythema, Purulent exudate, Separation of
deep tissues, Isolation of bacteria, and duration of inpatient Stay 4 ) scores were also significantly
better with high levels of oxygen (3 vs. 5, p=0.01). Although longer follow-up might have
identified additional wound infections, the authors argue that it was unlikely that these events
would take place preferentially in one group as the proposed therapeutic effect of oxygen
appears limited to the immediate perioperative period. 3 Admission to the intensive care unit and
death were less frequent in the intervention group, but the difference failed to achieve statistical
significance.
Two additional randomized controlled trials of perioperative supplemental oxygen were
identified. 5,6 Both found a significant reduction in postoperative nausea and vomiting, but neither
study evaluated the effect on wound infections.
Potential for Harm
The study by Greif et al reported no adverse affects related to the intervention. Several
potential risks of high oxygen concentrations should be noted. High oxygen concentrations may
present a fire hazard when heated surgical instruments (eg, lasers) are introduced into the
airway. 7-11 Such concentrations can also induce lung injury in certain vulnerable patients 12 or
precipitate atelectasis in patients at risk. 3,13,14 Hyperoxic mixtures may increase oxidative
myocardial injury in patients undergoing cardiopulmonary bypass. 15 Finally, patients who
undergo resuscitation with 100% oxygen may have worsened neurologic outcomes, possibly also
as a result of increased oxygen free-radical generation. 16,17
Costs and Implementation
The incremental direct costs associated with administering high oxygen concentrations
are minimal, as oxygen delivery systems are elements of routine perioperative care and employ
equipment readily available in operating rooms.
235

Comment
Administration of perioperative oxygen in high concentrations seems a promising
adjunctive therapy: the practice is simple, the equipment needed is readily available, and a
multicenter randomized trial has demonstrated its efficacy.
However, there are significant questions about the generalizability of the approach to
expanded populations of surgical patients. All patients in the Grief et al study had core
temperature maintained at 36ºC, were aggressively hydrated, and had postoperative pain treated
with opioids in order to maximize wound perfusion. To what degree the effectiveness of the
practice is affected by changes in these “co-interventions” has not been assessed. There is reason
for concern regarding use of high concentrations of oxygen in patients undergoing procedures
associated with low blood flow (eg, cardiopulmonary bypass), or in whom local production of
oxygen free radicals may cause further organ injury (eg, patients with head trauma).
Additionally, questions remain regarding whether modifications to the protocol used
would impart similar or greater benefit. For example, would oxygen administration by nasal
cannula at 10 LPM be as effective as oxygen delivered by a sealed mask? Would longer duration
of therapy impart additional benefit? These questions should be answered in future trials.
Table 20.3.1. Randomized controlled trial of supplemental perioperative oxygen*
Study Study Population Intervention Results†
Greif,
2000 3
500 patients
undergoing colorectal
resection; multicenter
study, 1996-98
80% oxygen, 20%
nitrogen during surgery
and the first 2 hours of
recovery
236
Wound infection:
ARR 0.06 (95% CI, 0.018-0.102)
RR 0.46 (95% CI, 0.25-0.88)
ASEPSIS § score: 3 vs. 5 (p=0.01)
ICU admission: 2.0% vs. 4.8%
(p=0.14)
Mortality: 0.4% vs. 2.4% (p=0.13)
* ARR indicates absolute risk reduction; CI, confidence interval; ICU, intensive care unit; and
RR, relative risk. The ASEPSIS scoring system incorporates Additional treatment, Serous
discharge, Erythema, Purulent exudate, Separation of deep tissues, Isolation of bacteria, and
duration of inpatient Stay 4 .
† Outcomes within 15 days of surgery, expressed as rates in intervention vs. control groups.
References
1. Hopf HW, Hunt TK, West JM, et al. Wound tissue oxygen tension predicts the risk of
wound infection in surgical patients. Arch Surg. 1997;132:997-1004.
2. Hopf H, Sessler DI. Routine postoperative oxygen supplementation. Anesth Analg.
1994;79:615-616.
3. Greif R, Akca O, Horn EP, Kurz A, Sessler DI. Supplemental perioperative oxygen to
reduce the incidence of surgical- wound infection. Outcomes Research Group. N Engl J
Med. 2000;342:161-167.
4. Wilson AP, Treasure T, Sturridge MF, Gruneberg RN. A scoring method (ASEPSIS) for
postoperative wound infections for use in clinical trials of antibiotic prophylaxis. Lancet.
1986;1:311-313.

5. Greif R, Laciny S, Rapf B, Hickle RS, Sessler DI. Supplemental oxygen reduces the
incidence of postoperative nausea and vomiting. Anesthesiology 1999;91:1246-1252.
6. Goll V, Akca O, Greif R, et al. Ondansetron is no more effective than supplemental
intraoperative oxygen for prevention of postoperative nausea and vomiting. Anesth Analg.
2001;92:112-117.
7. Healy GB. Complications of laser surgery. Otolaryngol Clin North Am. 1983;16:815-820.
8. Hunsaker DH. Anesthesia for microlaryngeal surgery: the case for subglottic jet
ventilation. Laryngoscope. 1994;104:1-30.
9. Aly A, McIlwain M, Duncavage JA. Electrosurgery-induced endotracheal tube ignition
during tracheotomy. Ann Otol Rhinol Laryngol. 1991;100:31-33.
10. Barnes AM, Frantz RA. Do oxygen-enriched atmospheres exist beneath surgical drapes
and contribute to fire hazard potential in the operating room? Aana J. 2000;68:153-161.
11. de Richemond AL, Bruley ME. Use of supplemental oxygen during surgery is not risk free.
Anesthesiology. 2000;93:583-584.
12. Donat SM, Levy DA. Bleomycin associated pulmonary toxicity: is perioperative oxygen
restriction necessary? J Urol. 1998;160:1347-1352.
13. Akca O, Podolsky A, Eisenhuber E, et al. Comparable postoperative pulmonary atelectasis
in patients given 30% or 80% oxygen during and 2 hours after colon resection.
Anesthesiology. 1999;91:991-998.
14. Lentschener C. Prevention of atelectasis during general anaesthesia. Lancet. 1995;346:514-
515.
15. Ihnken K, Winkler A, Schlensak C, et al. Normoxic cardiopulmonary bypass reduces
oxidative myocardial damage and nitric oxide during cardiac operations in the adult. J
Thorac Cardiovasc Surg. 1998;116:327-334.
16. Zwemer CF, Whitesall SE, D'Alecy LG. Cardiopulmonary-cerebral resuscitation with
100% oxygen exacerbates neurological dysfunction following nine minutes of
normothermic cardiac arrest in dogs. Resuscitation. 1994;27:159-170.
17. Rubertsson S, Karlsson T, Wiklund L. Systemic oxygen uptake during experimental
closed-chest cardiopulmonary resuscitation using air or pure oxygen ventilation. Acta
Anaesthesiol Scand. 1998;42:32-38.
Subchapter 20.4. Perioperative Glucose Control
Background
Diabetes is a well-known risk factor for perioperative medical complications. Poor
glucose control is an independent risk factor for surgical site infections 1-5 in a range of surgical
procedures. Increased risk for infection is thought to result from a combination of clinically
apparent effects of longstanding hyperglycemia (eg, macro- and microvascular occlusive
disease) and subtle immunologic defects, most notably neutrophil dysfunction. 6-12
Hyperglycemia may also impair the function of complement and antibodies, reducing the
opsonic potential of these factors and impairing phagocytosis, further reducing barriers to
infection. 13,14 Although many of the clinically apparent manifestations of diabetes are not easily
reversed in the perioperative period, there is a small literature that suggests that improving
glucose control can improve immunologic function and reduce the incidence of surgical site
infections (SSI). 6-8,12
237

Perioperative management of glucose for diabetic patients commonly includes
withholding or administering a reduced dose of the patients’ usual hypoglycemic agent(s) and
commencing a low-rate intravenous glucose infusion while patients are NPO prior to surgery.
The infusion is continued postoperatively until the patient is able to eat and resume outpatient
diabetes therapy. Often a sliding scale insulin regimen, a schedule of subcutaneous regular
insulin dosage contingent on capillary blood glucose measurements, is also continued through
the perioperative period. However, use of a sliding scale may result in wide variations in serum
glucose, 15 opening the rationale of this method to question. 16-18
Practice description
Aggressive glucose control in the perioperative period can be achieved using a
continuous intravenous insulin infusion (CII). Nursing staff monitor fingerstick (or arterial line
drop-of-blood sample) glucose measurements and adjust the infusion rate based on a protocol
intended to maintain serum glucose within a certain range. For example, the target range for the
original Portland Protocol was between 151 and 200 mg/dL. 16,19,20 In the most recent version, the
range is between 125 and 175 mg/dL. 21
Prevalence and Severity of the Target Safety Problem
Little evidence exists to describe the practice of CII in prevention of surgical site
infections in broad surgical practice. The small amount of evidence available describes its use in
patients undergoing cardiac surgery, primarily coronary artery bypass grafting (CABG).
Diabetes is a well-described risk factor for sternal wound infections, a catastrophic complication
of median sternotomy. 19,22-25 Sternal wound infections occur in 0.8% to 2% of unselected
patients undergoing median sternotomy and CABG. 20,22,23 Diabetic patients, who comprise
between 17 and 20% of all patients undergoing CABG, have been reported to have an incidence
of sternal wound infections as high as 5.6%. 26 Such infections are associated with marked
increases in morbidity and costs. Furnary et al reported that patients with sternal wound
infections had an average increased length of stay of 16 days and a higher mortality rate (19%
vs. 3.8% in patients without sternal wound infections). 20 (See also Subchapter 20.1).
Opportunities for Impact
More than 700,000 Americans underwent open-heart surgery in 1998 alone. 27 Up to 20%
of these patients may be candidates for continuous insulin infusion. Although CII is included in
the recent ACC/AHA Guidelines for CABG Surgery, 28 there are no data on the extent to which
the measure is currently used during cardiac or other surgical procedures.
Study Designs and Outcomes
We identified one prospective before-after study that compared rates of deep sternal
wound infections (DSWI) in diabetic patients undergoing CABG before and after
implementation of an aggressive CII protocol. 20 DSWI included infections involving the sternum
or mediastinal tissues, including mediastinitis. An older study from the same authors was not
reviewed as it reported findings at an earlier point in the same trial. 19 Additional studies
examined the use of CII in perioperative patients but did not report Level 1 clinical outcomes
relevant to patient safety (eg, mortality, wound infection) and were also not reviewed. 29
238

Evidence for Effectiveness of the Practice
Furnary et al found that aggressive glucose control with CII was associated with a
reduction in deep sternal wound infections. 20 The effect of the intervention remained statistically
significant in a logistic regression model adjusting for multiple potential confounding variables.
Furthermore, the demographic characteristics were generally biased against the CII group, which
had a significantly higher percentage of patients with hypertension, renal insufficiency, and
obesity but fewer patients with congestive heart failure. However, the authors did not adjust for
long-term markers of glucose control such as glycosylated hemoglobin, nor did they describe
other changes in patient care systems that resulted from changing patients to insulin infusions.
Continuous insulin infusions require closer attention by nursing staff both for monitoring of
infusion equipment and for frequent measurements of blood glucose. It is possible that the
improved outcomes were due to closer overall attention to the patient. Although 74% of DSWI
occurred after initial discharge (raising the concern that the shorter length of stay in the sliding
scale insulin group may have resulted in some infection not being detected), the authors reported
that they directly followed-up all diabetic patients for one year from the time of surgery. 30 The
personnel, equipment, surgical techniques, and use of prophylactic antibiotics were similar
throughout the study period. 31 Nonetheless, it is likely that secular trends in the care of patients
undergoing cardiac surgery account for some of the impact attributed to CII.
Potential for Harm
Hypoglycemic episodes are the most concerning adverse event associated with intensive
glucose management with intravenous insulin. These episodes result in a range of medical
complications, from delirium to myocardial infarction resulting from increased sympathetic
activity. Furnary noted that, using the standardized protocol in their study, no cases of
symptomatic hypoglycemia occurred in either group of patients. 30 However, CII protocols
intended to maintain normoglycemia in surgical patients have been associated with high rates
(40%) of postoperative hypoglycemia requiring treatment (

intervention group, the number of DSWIs prevented was 10 (95% CI: 4-21) and the average cost
to prevent one DSWI was approximately $21,000 (95% CI: $10,000-$52,500). Of course, these
figures do not incorporate the potential effects of the intervention on other sites of infection,
mortality, adverse events, and patients’ preferences (utilities) for these possible health states.
Comment
An increasing body of evidence demonstrates that tight control of blood glucose
improves overall outcomes of patients with diabetes. 35-37 Emerging data, coupled with an
increasing appreciation of the deleterious effects of hyperglycemia on immune function, strongly
support the supposition that aggressive control of perioperative glucose reduces the incidence of
surgical site infections. Although the practice has been implemented at a number of institutions
and is also being used in diabetic patients undergoing non-cardiac surgeries, 34 studies of its
effectiveness in these settings have not yet been published. Until additional evidence is available,
preferably from blinded randomized controlled trials, the intervention can be considered
promising but not yet proven to be causally associated with improved outcomes.
240

Table 20.4.1. Prospective, before-after study of aggressive perioperative glucose control*
Study Study
Population
Furnary,
1999 20
2467 diabetic
patients
undergoing
cardiac
surgery at a
community
hospital
Comparison Groups Results†
968 patients treated with
sliding scale SQ insulin
(1987-91)
1499 patients treated with
CII to target glucose of
150-200 mg/dL until
POD 3 (1991-97)
241
Deep surgical wound infections
Unadjusted: 1.9% vs. 0.8% (p=0.011)
Adjusted RR 0.34 (95% CI: 0.14-0.74)
Mortality: 6.1% vs. 3.0% (p=0.03)
Length of Stay: 10.7d vs. 8.5d (p

11. Mowat A, Baum J. Chemotaxis of polymorphonuclear leukocytes from patients with
diabetes mellitus. N Engl J Med. 1971;284:621-627.
12. MacRury SM, Gemmell CG, Paterson KR, MacCuish AC. Changes in phagocytic function
with glycaemic control in diabetic patients. J Clin Pathol. 1989;42:1143-1147.
13. Hennessey PJ, Black CT, Andrassy RJ. Nonenzymatic glycosylation of immunoglobulin G
impairs complement fixation. JPEN J Parenter Enteral Nutr. 1991;15:60-64.
14. Black CT, Hennessey PJ, Andrassy RJ. Short-term hyperglycemia depresses immunity
through nonenzymatic glycosylation of circulating immunoglobulin. J Trauma.
1990;30:830-832.
15. Queale WS, Seidler AJ, Brancati FL. Glycemic control and sliding scale insulin use in
medical inpatients with diabetes mellitus. Arch Intern Med. 1997;157:545-552.
16. Zinman B. Insulin regimens and strategies for IDDM. Diabetes Care. 1993;16(Suppl 3):24-
28.
17. Kletter GG. Sliding scale fallacy. Arch Intern Med. 1998;158:1472.
18. Shagan B. Does anyone here know how to make insulin work backwards? Why sliding
scale insulin coverage doesn't work. Pract Diabetol. 1990;9:1-4.
19. Zerr KJ, Furnary AP, Grunkemeier GL, Bookin S, Kanhere V, Starr A. Glucose control
lowers the risk of wound infection in diabetics after open heart operations. Ann Thorac
Surg. 1997;63:356-361.
20. Furnary AP, Zerr KJ, Grunkemeier GL, Starr A. Continuous intravenous insulin infusion
reduces the incidence of deep sternal wound infection in diabetic patients after cardiac
surgical procedures. Ann Thorac Surg. 1999;67:352-360.
21. Starr Wood Cardiac Group. Starr Wood Research - Continuous Intravenous Insulin
Infusion. Available at: http://www.starrwood.com/research/insulin.html. Accessed June 11,
2001.
22. Borger MA, Rao V, Weisel RD, Ivanov J, Cohen G, Scully HE, et al. Deep sternal wound
infection: risk factors and outcomes. Ann Thorac Surg. 1998;65:1050-1056.
23. Trick WE, Scheckler WE, Tokars JI, Jones KC, Reppen ML, Smith EM, et al. Modifiable
risk factors associated with deep sternal site infection after coronary artery bypass grafting.
J Thorac Cardiovasc Surg. 2000;119:108-114.
24. Zacharias A, Habib RH. Factors predisposing to median sternotomy complications. Deep
vs superficial infection. Chest. 1996;110:1173-1178.
25. Whang W, Bigger JT. Diabetes and outcomes of coronary artery bypass graft surgery in
patients with severe left ventricular dysfunction: results from The CABG Patch Trial
database. The CABG Patch Trial Investigators and Coordinators. J Am Coll Cardiol.
2000;36:1166-1172.
26. Golden SH, Peart-Vigilance C, Kao WH, Brancati FL. Perioperative glycemic control and
the risk of infectious complications in a cohort of adults with diabetes. Diabetes Care.
1999;22:1408-1414.
27. American Heart Association. 2001 Heart and Stroke Statistical Update. Available at:
http://www.americanheart.org/statistics/medical.html. Accessed June 11, 2001.
28. Eagle KA, Guyton RA, Davidoff R, Ewy GA, Fonger J, Gardner TJ, et al. ACC/AHA
Guidelines for Coronary Artery Bypass Graft Surgery: A Report of the American College
of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee
to Revise the 1991 Guidelines for Coronary Artery Bypass Graft Surgery). American
College of Cardiology/American Heart Association. J Am Coll Cardiol. 1999;34:1262-
1347.
242

29. Lazar HL, Chipkin S, Philippides G, Bao Y, Apstein C. Glucose-insulin-potassium
solutions improve outcomes in diabetics who have coronary artery operations. Ann Thorac
Surg. 2000;70:145-150.
30. Furnary AP. Continuous intravenous insulin infusion reduces the incidence of deep sternal
wound infection in diabetic patients after cardiac surgical procedures [discussion]. Ann
Thorac Surg. 1999;67:360-62.
31. CTSNet, the Cardiothoracic Surgery Network. Use of historical controls may weaken
conclusions. Available at: Available at: http:www.ctsnet.org/forum/78/0/2240. Accessed
June 18, 2001.
32. Chaney MA, Nikolov MP, Blakeman BP, Bakhos M. Attempting to maintain
normoglycemia during cardiopulmonary bypass with insulin may initiate postoperative
hypoglycemia. Anesth Analg. 1999;89:1091-1095.
33. Star Wood Cardiac Group. The Portland Insulin Protocol - Frequently Asked Questions
Page 1. Available at: http://www.starrwood.com/research/Insulin_FAQ1.html. Accessed
June 11, 2001.
34. Starr Wood Cardiac Group. The Portland Insulin Protocol - Frequently Asked Questions
Page 2. Available at: http://www.starrwood.com/research/Insulin_FAQ2.html. Accessed
June 11, 2001.
35. The effect of intensive treatment of diabetes on the development and progression of longterm
complications in insulin-dependent diabetes mellitus. The Diabetes Control and
Complications Trial Research Group. N Engl J Med. 1993;329:977-986.
36. The effect of intensive diabetes therapy on the development and progression of neuropathy.
The Diabetes Control and Complications Trial Research Group. Ann Intern Med.
1995;122:561-568.
37. Retinopathy and nephropathy in patients with type 1 diabetes four years after a trial of
intensive therapy. The Diabetes Control and Complications Trial/Epidemiology of
Diabetes Interventions and Complications Research Group. N Engl J Med. 2000;342:381-
389.
243

244

Chapter 21. Ultrasound Guidance of Central Vein Catheterization
Jeffrey M. Rothschild, MD, MPH
Harvard Medical School
Background
The multiple indications for central venous catheters (CVCs) include parenteral nutrition,
intravascular depletion, access for vasoactive medications, hemodynamic monitoring,
cardiopulmonary arrest, difficult peripheral intravenous (IV) access and long-term IV access for
medications, such as antibiotics. 1,2 While these catheters can be life saving, they are also
associated with significant risks. 3 These risks increase in association with several characteristics,
including patient anatomy (eg, morbid obesity, cachexia, or local scarring from surgery or
radiation treatment), patient setting (eg, patients receiving mechanical ventilation or during
emergencies such as cardiac arrest) and co-morbidities (eg, bullous emphysema or
coagulopathy). 3-5
CVCs are placed by clinicians whose training and experience may vary greatly. The
procedure takes place in a variety of hospital settings including intensive care units, emergency
departments, operating rooms, pre- and post-anesthesia care units, hemodialysis units, cardiac
catheterization units and other inpatient settings. Outpatient placement of CVCs has also become
commonplace, occurring in hemodialysis centers and oncology centers providing outpatient
chemotherapy.
Percutaneous insertions of CVCs are usually performed by “blind” techniques that rely
on anatomic landmarks – ie, palpable or visible structures with known relationships to the
desired vein. For example, the infraclavicular approach to the subclavian vein requires correct
localization of the clavicle reference site, suprasternal notch and sternocleidomastoid-clavicular
triangle landmarks, proper positioning of the patient and operator and correct venipuncture point
depth, direction and insertion angle. Analogously, the various approaches to the internal jugular
vein require thorough knowledge of this vein’s course in relation to the sternocleidomastoid
muscle and carotid artery. 1
Newer technologies, such as portable ultrasound (US) devices, provide bedside imaging
of the central veins during catheter placement. The advantages associated with US guided CVC
placement include detection of anatomic variations and exact vessel location, avoidance of
central veins with pre-existing thrombosis that may prevent successful CVC placement, and
guidance of both guidewire and catheter placement after initial needle insertion. This review
assesses the impact of real-time ultrasound guidance on improving the safety of CVC insertions.
Practice Description
Real-time ultrasound guidance of CVC insertion provides the operator with visualization
of the desired vein and the surrounding anatomic structures prior to and during insertion of the
needle, guidewire and catheter. Previous studies of US location of vessels followed by
subsequent catheter placement with landmark techniques found no advantages over standard
landmark techniques. 3 Real-time US guidance, on the other hand, appears to improve the success
rate and decrease the complication rate associated with CVC placement.
Two types of real-time ultrasound guidance are described. The Doppler US guidance
method include audio-guided Doppler, 6 fingertip pulsed Doppler 7 and probe-in-the-needle
245

technology. 6,8-10 The non-Doppler US guidance methods (subsequently referred to as US
guidance) includes US with needle guidance 11,12 or without needle guidance. 13-15
Prevalence and Severity of the Target Safety Problem
The annual number of all CVCs insertions in the United States is not known, but is
estimated at “several million” for subclavian-vein catheters. 3 When aggregated with the various
types of catheters placed into the central venous circulation and the increasing utilization of
CVCs among routine surgical patients, critically ill patients (in both emergency departments and
intensive care units), and for the management of many patients undergoing hemodialysis or
chemotherapy, the total number of CVC placements may be many times greater than estimates
for subclavian CVCs alone. 3
Unsuccessful insertion of CVCs may occur in up to 20% of cases. 3-5 The hazards
associated with attempted CVC insertion (whether successful or not) include arterial puncture,
hematoma, pneumothorax, hemothorax, chylothorax, brachial plexus injury, arrhythmias, air
embolus, catheter malposition and catheter knotting. 3-5 Other complications associated with
CVCs, such as infection, thrombosis, arterial-venous fistula and vascular or cardiac erosion, are
not usually associated with needle insertion but occur after catheter placement. 3-5
The frequency of complications associated with CVC placement is quite variable, largely
due to differences among selected venous insertion sites, the degree of prior operator experience
and the presence of previously described risk factors. In general, the rate of major CVC
complications (eg, pneumothorax or vessel laceration requiring repair) and minor complications
(eg, arterial puncture without significant hemorrhage, transient catheter malposition) is between
0.5 and 10%. 3-5
In addition to complications, several quality-of-care issues are associated with problems
in CVC insertion. For example, a CVC insertion that requires multiple attempts may engender
considerable patient anxiety and discomfort. More importantly, a prolonged insertion process
may delay the infusion of life-saving fluids or medications during an emergency.
Opportunities for Impact
The majority of CVC insertions are placed using the landmark method. As set forth
above, the number of catheters placed annually and the proportion currently inserted without US
guidance is not known. Also unknown is the proportion of catheters placed in those patients who
may benefit most from the US technique—those with multiple risk factors, those in high risk
settings such as the intensive care unit, or those undergoing catheter placement by inexperienced
operators.
There are a variety of catheters that require access to the central veins. These include
single and multi-lumen CVCs, tunneled and non-tunneled catheters, and larger, more rigid
catheter introducers that permit passage of thinner, more flexible devices (eg, pulmonary artery
catheters, diagnostic cardiac catheters and temporary transvenous pacemakers). All centrallyplaced
catheters require an initial needle insertion into the vein, followed by a guidewire to
permit passage of the catheter.
Study Designs
One meta-analysis and 10 original studies were analyzed. Among the 10 original studies,
9 were randomized control studies and 1 study was a quasi-randomized control trial (see Table
21.1). 11 The meta-analysis 16 includes 6 references cited in this chapter. 6,8,10,12-14 Four articles
cited in this chapter were not cited in the 1996 meta-analysis, including 3 that were published
246

after 1996 9,15,17 and one that included a quasi-randomized design. 17 Two studies included in the
meta-analysis were not included in this chapter because they were from non-English language
journals. 7,18
All of the studies analyzed for this chapter were prospective, non-blinded and
randomized, with the exception of a quasi-randomized design involving alternate week
allocation of patients to receive the intervention. 11 Randomization was at the patient (rather than
physician) level in all studies.
The 10 original studies include 5 using an ultrasound guidance technique without
Doppler and 5 using the Doppler US technique. Sites of catheterization included the internal
jugular (IJ) veins in 6 studies, the subclavian (SC) veins in 3 studies and the femoral veins in 1
study. The study populations are diverse and include intensive care unit patients, surgical
patients both preoperatively and intraoperatively, cardiac patients in both the catheterization and
coronary care units, emergency department patients in cardiopulmonary arrest and oncology
center outpatients.
Examples of studies excluded from analysis are studies lacking comparison control
groups, 19-27 studies of US use but without real-time guidance, 3,28 and simulations of CVC
placement rather than completed procedures. 29
Study Outcomes
Studies included for review reported a combination of clinical outcomes. These outcomes
include Level 1 complications that resulted in increased patient morbidity (eg, pneumothorax)
and Level 2 complications that represent potential adverse events (eg, unwanted arterial puncture
without sequelae). Most studies also reported the number of venipuncture attempts to achieve
successful CVC placement as well as the time required to complete successful CVC insertions.
These are considered Level 2 outcomes because increased venipuncture attempts are associated
with increased complication rates. 4
Evidence for Effectiveness of the Practice
The 1996 meta-analysis 16 estimated that real-time US guidance for CVC insertion is
associated with a significant reduction in placement failures as compared with the usual
landmark techniques (relative risk 0.32, 95% CI: 0.18-0.55). In addition, this review estimated
that US guidance results in decreased complications during attempted CVC placements (relative
risk 0.22, 95% CI: 0.10-0.45), corresponding to a relative risk reduction of 78%. 16 The mean
number of attempted venipunctures till successful CVC insertion was significantly reduced with
real-time US guidance (relative risk 0.60, 95% CI: 0.45-0.79), corresponding to a relative risk
reduction of 40%. 16
Two of the 3 studies reviewed for this chapter and not included in the meta-analysis
(because they were published subsequently) deserve mention because of contrary findings. Both
1998 studies 9,17 included operators with significant prior experience placing CVCs by the usual
landmark method. The overall failure rate for both landmark and US guidance techniques was
very low in one study. 17 The other study found statistically insignificant negative results
associated with US guidance, owing to a high failure rate during the initial learning period for
the newly-introduced US guidance technique, and a very low (1.3%) overall complication rate
for CVC placement. 9
With the exception of femoral venous catheterization during cardiopulmonary
resuscitation, 15 the studies reviewed for this chapter did not find reductions in insertion time
when using real-time US guidance. With 2 exceptions, 910 the cited studies reached statistical
247

significance for at least one of the 3 outcome measures (morbidity, potential adverse events, and
number of venipuncture attempts). The most favorable outcomes associated with real-time US
guidance were found in studies of inexperienced operators. 6,12,13
Potential for Harm
The additional equipment and manipulation associated with real-time US guidance for
CVC insertion may increase the rate of catheter-related infections, but published studies have not
included these complications. In emergency settings, the increased length of time required to
place CVCs under US guidance (usually an additional 30 seconds to several minutes) may result
in unacceptable delays. Potential harmful consequences resulting from real-time US guidance for
CVC placement relate to changes in training and subsequent dependence on this technology.
Supporters of this technology argue that increased competence and anatomic knowledge gained
with US guidance will enhance performance of customary, unaided CVC placement. It is unclear
if trainees who have performed CVC placement only with US assistance will have different
complication rates when placing CVCs in practice settings without US equipment. Therefore, for
certification of qualification, trainees may need to demonstrate competence with both US and
non-US guided CVC placement.
Costs and Implementation
The major impediments to the widespread implementation of US guidance for CVC
insertion are the purchase costs of the US machines. A typical machine costs $11,000-16,000
(including the probes), with a single machine able to serve most critical care units. Depending on
the layout of units placing CVCs, an average 400-bed hospital would require 1-3 machines for
use outside of the operating room. (Departments of Anesthesia may require only one or two
machines, as experienced anesthesiologists can continue to place most CVCs without US
guidance). Hospitals in which nurses place peripherally inserted central catheter (PICC) lines
using US guidance typically facilitate this function with a single machine, which could be dually
used for CVCs depending on workflow and volume. In one study, the price of the Doppler-Smart
needles (which are not required for non-Doppler US guidance) was $40-70 as compared with $3-
5 for the standard needles. 9 The cost of training new operators (including those whose only prior
experience is with the landmark technique) requires further evaluation.
248

Comment
Real-time US guidance for CVC insertion, with or without Doppler assistance, improves
catheter insertion success rates, reduces the number of venipuncture attempts prior to successful
placement, and reduces the number of complications associated with catheter insertion.
However, these benefits may not accrue until after the initial learning period for operators
already experienced in the landmark techniques. 9
There are no studies comparing the impact of CVC insertion with US guidance on overall
patient outcomes (eg, mortality, length of stay). In addition, many of the complications
associated with CVC insertion are minor or easily treated. The reduction in venipuncture
attempts is likely associated with reductions in the pain and discomfort associated with CVC
placement, though this has not been measured.
The greatest benefit of US guidance may apply to the novice or inexperienced operator
and for all operators in high-risk situations. Patients with one or more risk factors, (eg, critically
ill patients on positive pressure ventilation with generalized edema and coagulopathy), may reap
the greatest benefit. CVC insertion training incorporating real-time ultrasound guided techniques
may provide additional valuable learning benefits for new operators. This knowledge may
improve the success rate of insertion of CVCs without US guidance. Simulation training has
demonstrated improved identification of the desired veins with US as compared to landmark
techniques. 29
Finally, it should be noted that in addition to real-time US guidance, other approaches
may reduce the risks associated with CVC insertion. PICC lines are gaining widespread
acceptance and may be an acceptable substitute for CVCs for certain indications (eg, long-term
IV access or parenteral nutrition). 30,31 US guidance has also been demonstrated to improve the
insertion of PICCs. 32-34 Increases in the use of PICCs may help justify the purchase of ultrasound
machines by individual hospitals.
For patients requiring replacement of existing CVCs, guidewire exchanges offer a way of
inserting new catheters without new venipuncture attempts. A systematic review has found that
guidewire exchanges are associated with fewer mechanical complications than new-site
replacement, but may be associated with greater risks for catheter-related infections (Chapter
16). 35
Alternative methods for teaching CVC insertion skills to novices (eg, first-year resident
physicians and medical students) have successfully employed multidisciplinary approaches
including cadaver demonstrations. 36 Future methods for teaching CVC insertion may employ
computerized technologies for simulations (see also Chapter 45). Haptic, or touch-related
techniques use virtual reality models to create immersive simulated environments that recreate
the sensation of performing a procedure. 37,38 Through the use of this and other new technologies,
novice operators may gain experience and confidence prior to clinical CVC insertion attempts
and further improve patient safety.
249

Table 21.1. Ultrasound and Doppler ultrasound guidance of central vein catheters*
Study Setting and
Population
Tertiary care, teaching
hospital ICU 13
Tertiary care, teaching
hospital, CT surgical
patients 14
Tertiary care, teaching
hospital, cardiac
patients 11
Urban, teaching
hospital ICU 12
Urban, teaching
hospital ED, during
CPR 15
Tertiary care, teaching
hospital, CT/ vascular
surgery patients 8
British hospital,
cardiac surgery and
ICU patients 10
Tertiary care, teaching
hospital ICU and OR;
high-risk patients 6
French teaching
hospital ICU; low-risk
patients 17
Tertiary care,
outpatient oncology
center 9
Year
Published
Intervention Study
Design,
Outcomes
Ultrasound
1990 US guidance for IJ CVC insertion
without needle guide; concurrent
feedback from an US technician
1991 US guidance (7.5 and 5.0 MHz
transducers) for IJ CVC insertion
without needle guide
1993 US guidance (7.5 MHz
transducer) of IJ cannulation for
cardiac catheterization and CVC
insertion, with needle guide
1995 US guidance (7.5 MHz
transducer) for SC CVC insertion
with needle guide
1997 US guidance (7.5 MHz
transducer) for femoral CVC
insertion without needle guide
Doppler Ultrasound
1994 Doppler US guidance of IJ CVC
insertion with probe in the needle
1994 Doppler US guidance of IJ CVC
insertion with probe in the needle
1995 Audio-guided Doppler US
guidance for IJ CVC insertion
with probe in the needle
1998 Pulsed Doppler US guidance for
SC CVC insertion without needle
guide
1998 Doppler US guidance for SC
CVC insertion with probe in the
needle
250
Level 1
Level 2
Level 1
Level 2
Level 1
Level 2
Level 1
Level 2
Level 1
Level 2
Level 1
Level 2
Level 1
Level 2
Level 1
Level 2
Level 1
Level 2
Level 1
Level 2
Relative Risk Reduction (%)†
Failed
Catheter
Insertion
100 NS
Mean
Insertion
Attempts
Required§
44 NA
Complications
100 44 83 NS
100 48 80
86 48 90
71 54 100
0 52 0
-50 NS
17 NS
63 18 NS
-32 §
-46 §
0
88
0 67
NA -53 NS
* CPR indicates cardiopulmonary resuscitation; CT, cardiothoracic; ED, emergency department; ICU, intensive care
unit; IJ, internaljugular vein; NA, not available; NS, not statistically significant; OR, operating room; RCT,
randomized controlled trial; SC, subclavian vein; and US, ultrasound guidance.
† The percentage relative risk reduction reflects the risks associated with CVC placement and the percentage change
resulting from US guidance. Negative values indicate an increase in risks associated with US guidance.
§ Relative risk reduction of the insertion attempts to success reflects the relative reduction in the mean number of needle
insertions attempted per patient until successful CVC placement resulting from US guidance.

References
1. Civetta JM, Taylor RW, Kirby RR. Critical Care. 3rd ed. Philadelphia: Lippincott-Raven
Publishers, 1997.
2. Shoemaker W, Ayres S, Grenvik A, Holbrook P. Textbook of Critical Care. 4 ed.
Philadelphia: WB Saunders, 2000.
3. Mansfield PF, Hohn DC, Fornage BD, Gregurich MA, Ota DM. Complications and failures
of subclavian-vein catheterization. N Engl J Med. 1994;331:1735-1738.
4. Sznajder JI, Zveibil FR, Bitterman H, Weiner P, Bursztein S. Central vein catheterization.
Failure and complication rates by three percutaneous approaches. Arch Intern Med.
1986;146:259-261.
5. Bernard RW, Stahl WM. Subclavian vein catheterizations: a prospective study. I. Noninfectious
complications. Ann Surg. 1971;173:184-190.
6. Gilbert TB, Seneff MG, Becker RB. Facilitation of internal jugular venous cannulation
using an audio-guided Doppler ultrasound vascular access device: results from a
prospective, dual-center, randomized, crossover clinical study. Crit Care Med. 1995;23:60-
65.
7. Branger B, Zabadani B, Vecina F, Juan JM, Dauzat M. Continuous guidance for venous
punctures using a new pulsed Doppler probe: efficiency, safety. Nephrologie. 1994;15:137-
140.
8. Gratz I, Afshar M, Kidwell P, Weiman DS, Shariff HM. Doppler-guided cannulation of the
internal jugular vein: a prospective, randomized trial. J Clin Monit. 1994;10:185-188.
9. Bold RJ, Winchester DJ, Madary AR, Gregurich MA, Mansfield PF. Prospective,
randomized trial of Doppler-assisted subclavian vein catheterization. Arch Surg.
1998;133:1089-1093.
10. Vucevic M, Tehan B, Gamlin F, Berridge JC, Boylan M. The SMART needle. A new
Doppler ultrasound-guided vascular access needle. Anaesthesia. 1994;49:889-891.
11. Denys BG, Uretsky BF, Reddy PS. Ultrasound-assisted cannulation of the internal jugular
vein. A prospective comparison to the external landmark-guided technique. Circulation.
1993;87:1557-1562.
12. Gualtieri E, Deppe SA, Sipperly ME, Thompson DR. Subclavian venous catheterization:
greater success rate for less experienced operators using ultrasound guidance. Crit Care
Med. 1995;23:692-697.
13. Mallory DL, McGee WT, Shawker TH, Brenner M, Bailey KR, Evans RG, et al.
Ultrasound guidance improves the success rate of internal jugular vein cannulation. A
prospective, randomized trial. Chest. 1990;98:157-160.
14. Troianos CA, Jobes DR, Ellison N. Ultrasound-guided cannulation of the internal jugular
vein. A prospective, randomized study. Anesth Analg. 1991;72:823-826.
15. Hilty WM, Hudson PA, Levitt MA, Hall JB. Real-time ultrasound-guided femoral vein
catheterization during cardiopulmonary resuscitation. Ann Emerg Med. 1997;29:331-336.
16. Randolph AG, Cook DJ, Gonzales CA, Pribble CG. Ultrasound guidance for placement of
central venous catheters: a meta-analysis of the literature. Crit Care Med. 1996;24:2053-
2058.
17. Lefrant JY, Cuvillon P, Benezet JF, Dauzat M, Peray P, Saissi G, et al. Pulsed Doppler
ultrasonography guidance for catheterization of the subclavian vein: a randomized study.
Anesthesiology. 1998;88:1195-1201.
251

18. Scherhag A, Klein A, Jantzen J. Cannulation of the internal jugular vein using 2 ultraosnic
techniques:A comparitive controlled study. Anaesthesist 1989;38:633-638.
19. Fry WR, Clagett GC, O'Rourke PT. Ultrasound-guided central venous access. Arch Surg.
1999;134:738-740.
20. Gordon AC, Saliken JC, Johns D, Owen R, Gray RR. US-guided puncture of the internal
jugular vein: complications and anatomic considerations. J Vasc Interv Radiol. 1998;9:333-
338.
21. Sato S, Ueno E, Toyooka H. Central venous access via the distal femoral vein using
ultrasound guidance. Anesthesiology. 1998;88:838-839.
22. Docktor B, So CB, Saliken JC, Gray RR. Ultrasound monitoring in cannulation of the
internal jugular vein: anatomic and technical considerations. Can Assoc Radiol J.
1996;47:195-201.
23. Hatfield A, Bodenham A. Portable ultrasound for difficult central venous access. Br J
Anaesth. 1999;82:822-826.
24. Gallieni M, Cozzolino M. Uncomplicated central vein catheterization of high risk patients
with real time ultrasound guidance. Int J Artif Organs. 1995;18:117-121.
25. Geddes CC, Walbaum D, Fox JG, Mactier RA. Insertion of internal jugular temporary
hemodialysis cannulae by direct ultrasound guidance—a prospective comparison of
experienced and inexperienced operators. Clin Nephrol. 1998;50:320-325.
26. Cavatorta F, Zollo A, Campisi S, Trabassi E, De Lucia E, Galli S, et al. Internal jugular
vein catheterization under echographic guidance. Int J Artif Organs. 1993;16:820-822.
27. Hrics P, Wilber S, Blanda MP, Gallo U. Ultrasound-assisted internal jugular vein
catheterization in the ED. Am J Emerg Med. 1998;16:401-403.
28. Hayashi H, Tsuzuku M, Amano M. Simplified echo-guided internal jugular vein puncture:
a comparison to the landmark-guided technique. Anesth.Analg. 1998;86:SCA89
29. Hirvela E, Parsa C, Aalmi O, Kelly E, Goldstein L. Skills and risk assessment of central
line placement using bedside simulation with 3-dimensional ultrasound guidance system.
Crit Care Med. 2000;28:A78-A78
30. Lam S, Scannell R, Roessler D, Smith MA. Peripherally inserted central catheters in an
acute-care hospital. Arch Intern Med. 1994;154:1833-1837.
31. Smith JR, MD, Friedell ML, Cheatham ML, Martin SP, Cohen MJ, et al. Peripherally
Inserted Central Catheters Revisited. Am J Surg. 1998;176:208-211.
32. LaRue GD,. Efficacy of Ultrasonography in Peripheral Venous Cannulation. J Intraven
Nurs. 2000; 23:29-34.
33. Sofocleous CT, Schur I, Cooper SG, Quintas JC, Brody L, Shelin R. Sonographically
guided placement of peripherally inserted central venous catheters: review of 355
procedures. AJR. 1998;170:1613-1616.
34. Chrisman HBM, Omary RAM, Nemcek AA, Jr, MD, Vogelzang RLM. Peripherally
Inserted Central Catheters: Guidance With Ultrasound Versus Venography in 2650
Patients. J Vasc Interv Radiol. 1998;9:173-174.
35. Cook D, Randolph A, Kernerman P, Cupido C, King D, Soukup C, et al. Central venous
catheter replacement strategies: a systematic review of the literature. Crit Care Med.
1997;25:1417-1424.
36. Nip IL, Haruno MM. A systematic approach to teaching insertion of a central venous line.
Acad Med. 2000;75:552.
37. Hiemenz L, Stredney D, Schmalbrock P. Development of the force-feedback model for an
epidural needle insertion simulator. Stud Health Technol Inform. 1998;50:272-277.
252

38. Kaufmann C, Rhee P, Burris D. Telepresence surgery system enhances medical student
surgery training. Stud Health Technol Inform. 1999;62:174-178.
253

254

Chapter 22. The Retained Surgical Sponge
Verna C. Gibbs, MD
Andrew D. Auerbach, MD, MPH
University of California, San Francisco School of Medicine
Background
Although less likely to garner public notoriety, errors relating to the failure to remove
surgical instruments at the end of a procedure, (ie, needles, knife blades, electrosurgical adapters
and safety pins) or sponges (known as gossypiboma; gossypium: Latin, cotton; boma: Swahili,
place of concealment) are no less egregious than the better known mishaps such as “wrong-site
surgery” (see Subchapter 43.2).
Retained materials may cause an acute foreign body reaction with local or systemic signs
that prompt investigation and reoperation. Alternatively, a fibrinous response may be elicited,
and the retained instrument or sponge may become apparent some time after the original surgical
procedure either serendipitously, or via fistulization into local structures. 1 The medical literature
is scattered with reports of presentations of retained sponges found days, months, or even years
after the original surgery. 2-5 While many cases of retained foreign body do not cause harm, some
clearly do. Nevertheless, the Joint Commission on Accreditation for Healthcare Organization’s
(JCAHO) sentinel event policy specifically mentions that “unintentionally retained foreign body
without major permanent loss of function” do not require reporting. 6 Although JCAHO’s
decision suggests that it considers these events less egregious than reportable sentinel events (eg,
wrong patient surgery), retained foreign body events are far more common. This chapter reviews
the problem and the scanty literature regarding safety practices to reduce the incidence of
retained sponges and instruments.
Practice Description
Surgeons and operating room teams rely upon the practice of sponge, sharp and
instrument counts as a means to eliminate retained surgical instruments. Counts are also a
method of infection control and inventory control, and a means to prevent injury from
contaminated sharps and instruments. Four separate counts have been recommended 7 : the first
when the instruments are set up or the sponges unpackaged, a second before the surgical
procedure begins, a third as closure begins, and the final count performed during subcuticular or
skin closure.
Use of this simple preventative measure is not universal. In fact, the process by which
counts are performed is not standardized and is often modified according to individual hospital
policy. Even when present, counts are frequently omitted or abbreviated in emergency or
transvaginal surgeries, or for vaginal deliveries. 8 An adjunctive procedure to the count, used
when the count could delay care and jeopardize patients’ lives or when an incorrect count is
established, is an x-ray examination to detect radiopaque objects. 1,7 Since this practice is not
routinely used it will not be discussed here.
255

Prevalence and Severity of the Target Safety Problem
A literature search revealed few data to describe population or even hospital-level
information regarding the prevalence of retained surgical materials. One study from a medical
malpractice insurance company reported 40 cases in a 7-year period, 9 or about 1% of all claims.
Because this estimate is based on malpractice insurance claims, it is sure to be a gross
underestimate of the actual incidence. A recent unstructured review cited “a prevalence ranging
from 1/100 to 1/5000,” and an associated mortality ranging from 11 to 35%, citing non-English
language medical references. 1 Other reports are based on case series or descriptions of unusual
presentations, as described above. Surgeons may not report these events for a variety of reasons,
not the least of which is fear of litigation
Opportunities for Impact
Without accurate prevalence information, the true magnitude of the opportunity for
impact is unclear.
Study Designs and Outcomes
Only one study provided even indirect evidence of the effectiveness of sponge and
instrument counts. Kaiser et al, using a retrospective review of medical malpractice claims data
from a statewide insurer in Massachusetts, reviewed 67 cases where retained sponges or surgical
materials were the primary reason for the claim. 9 This study is a case series without any controls
(Level 4 design, Level 2 outcomes) which reported only the outcome of retained sponges, rather
than the clinical consequences of these errors.
Evidence for Effectiveness of the Practice
The Kaiser et al study reported that 55% of retained sponges were found after abdominal
surgery and 16% after vaginal delivery. In cases with retained sponges, sponge counts had been
falsely correct in 76% of non-vaginal surgeries; in 10% no sponge count had been performed at
all. Falsely correct sponge counts were attributed to team fatigue, difficult operations, sponges
“sticking together,” or a poor counting system. Incorrect sponge counts that were accepted prior
to closure resulted from either surgeons’ dismissing the incorrect count without re-exploring the
wound, or nursing staff allowing an incorrect count to be accepted. Interestingly, in 3 of 29 cases
in which intraoperative x-rays were used to detect radiopaque sponges, the radiograph was
falsely negative. 9
Comment
Although literature describing the incidence of iatrogenic foreign bodies is highly limited
in quantity and quality, it is unlikely that these events are as rare as other iatrogenic
complications that have drawn considerable national attention. The existing system of sponge
and instrument counts probably works well, but we have no evidence to describe its actual
failure rate. The little existing evidence suggests that it fails due to human-related factors (ie, the
count is not performed, or is ignored, and that ancillary methods such as x-rays are also fallible.
Although some have advocated CT or ultrasonography as additional methods to reduce rates of
these adverse events, it is possible that other technologies (eg, inventory control devices used in
retail stores and libraries, possibly including bar coding (Subchapter 45.1)) may prove to be
useful adjuncts. However, there are obvious logistical challenges that make such technologies
too impractical at the present time. For now we are left with a paucity of data regarding the
256

prevalence of this error and the effectiveness of preventative measures. Use of anonymous
reporting systems may reduce the fear of litigation associated with iatrogenic foreign bodies, and
may allow for more accurate assessment of the incidence and causes of these events.
References
1. Lauwers PR, Van Hee RH. Intraperitoneal gossypibomas: the need to count sponges.
World J Surg. 2000;24:521-527.
2. Wig JD, Goenka MK, Suri S, Sudhakar PJ, Vaiphei K. Retained surgical sponge: an
unusual cause of intestinal obstruction. J Clin Gastroenterol. 1997;24:57-58.
3. Scott WW, Beall DP, Wheeler PS. The retained intrapericardial sponge: value of the lateral
chest radiograph. AJR Am J Roentgenol. 1998;171:595-597.
4. Sinha SK, Gupta S, Behra A, et al. Retained surgical sponge: an unusual cause of
malabsorption. Trop Gastroenterol. 1999;20:42-44.
5. Burrel M, Capurro S, Arguis P, Vilana R. Sonographic appearance of a retained surgical
sponge in the neck. J Clin Ultrasound. 2000;28:311-313.
6. The Joint Commission on Accreditation of Healthcare Organizations. Sentinel Event Alert.
Accreditation Committee Approves Examples Of Voluntarily Reportable Sentinel Events.
May 11, 1998. Available at: http://www.jcaho.org/edu_pub/sealert/sea4.html. Accessed
June 23, 2001.
7. Recommended practices for sponge, sharp, and instrument counts. AORN Recommended
Practices Committee. Association of periOperative Registered Nurses. AORN J.
1999;70:1083-9.
8. Horty J. Nurses, surgeon not negligent in retained instrument suit. Kissinger v Turner. OR
Manager. 1995;11:26, 30.
9. Kaiser CW, Friedman S, Spurling KP, Slowick T, Kaiser HA. The retained surgical
sponge. Ann Surg. 1996;224:79-84.
257

258

Chapter 23. Pre-Anesthesia Checklists To Improve Patient Safety
Andrew D. Auerbach, MD, MPH
University of California, San Francisco School of Medicine
Harvey J. Murff, MD
Harvard Medical School
Salim D. Islam, MD
University of California, San Francisco School of Medicine
Background
No matter how rote the task or how vigilant the anesthesiologist, “slips” and other errors
represent expected aspects of human performance. 1 Evaluation and subsequent improvement of
standard checkout procedures promises to increase patient safety in the perioperative period by
removing more of the “human factors” so often implicated in anesthesia adverse events. 2 The use
of pre-flight checklists has been considered a key method in improving airline safety, largely due
to the regular systematizing of complex procedures, and improvement of team dynamics through
authority-neutral tasks. A checklist system has been proposed as part of routine pre-anesthesia
care, with the American Society of Anesthesiologists and the US Food and Drug Administration
(FDA) issuing general guidelines supporting checklists in 1986. 3 Subsequently, anesthesia
professional societies in Great Britain and Europe adopted similar standards. 4
Practice Description
In 1987, the FDA published the “Anesthesia Apparatus Checkout Recommendations” in
the Federal Register (February, 1987). This “checkout list” provides practitioners with a
standardized approach to checking anesthetic equipment prior to its use in order to ensure that
the delivery system is correctly connected, adjusted, and functional. The original checklist
included 24 specific processes to be performed as an initial checkout at the beginning of each
day; 11 are performed between cases and after initial equipment evaluation. 5 Many clinicians
regarded these protocols as too long and complex for routine use. Parties involved in revising the
recommendations agreed that the average clinician should be able to check an anesthesia
machine in 5 minutes or less (not possible with the 1987 recommendations). The revised
checklist included only 14 major processes, 9 of which can be omitted or substantially
abbreviated when the anesthesia provider uses the same equipment in successive cases. 6 The
revised recommendations are available online. 7
Prevalence and Severity of the Target Safety Problem
The earliest observational studies of mishaps within anesthesia found that equipment
failure was the cause of 14% of anesthesia critical incidents. 8 Subsequent studies reduced this
estimate considerably. Although equipment failure is now implicated in only 4% of anesthesia
adverse events, 22% are related to failing to check equipment adequately. 2 Equipment failures
can result in delivery of hypoxic gas mixtures or excessive doses of inhalational agent, or
hypoventilation due to ventilator failure. These situations can be catastrophic if unrecognized,
and even if recognized may result in significant morbidity (eg, delayed extubation, stroke, or
myocardial infarction). 2,9
259

Opportunities for Impact
The FDA checkout list is considered a template which local users are “encouraged to
modify to accommodate differences in equipment and variations in local clinical practice.”
Estimating the frequency with which it or other checklists are used (in modified or unmodified
forms) in anesthesia practice would be speculative, 5 but a survey of 4 states suggests that usage
is minimal. 3 There are 41.5 million inpatient procedures each year requiring anesthesia in the
US. Probably about half of these involve general anesthesia. Therefore, although the frequency
of equipment failure is low, any improvement in safety from anesthesia preoperative checklists
could have substantial impact.
Study Designs
Using a structured MEDLINE search, we identified 15 articles that discussed anesthesia
checklists. Of these, only 2 studies came close to meeting our inclusion criteria (Chapter 3). 3,10
Because no other studies could be found, we abstracted these studies and review them here
(Table 23.1).
The first investigation evaluated the ability of participating providers to detect
standardized equipment faults using their own checking methods compared with the FDA
checkout list. 3 Although the study involved a prospective design, it does not fit neatly into our
classification system because the participants (anesthesiology residents and practitioners) served
as their own controls. The second study 10 involved the revised FDA checkout list, and its design
was modeled on the previous study.
Study Outcomes
The first study’s outcome of interest was the detection of 4 standardized equipment faults
created by the investigators (Level 2). 3 The second study used similarly designed outcomes
(simulated equipment faults), but defined detection of at least 50% of these faults as the primary
outcome (Level 2). 10 Neither study directly connected the use of pre-anesthesia checklists to
patient outcomes, although inadequate preanesthesia equipment checks have been implicated in
adverse events related to equipment failures, as discussed above. 2
Evidence for Effectiveness of the Practice
March et al 3 found that the FDA checklist was more likely to detect faults with the
nitrous oxide system (65% vs. 43%, p

to anesthesia providers’ own methods, the observed utility of this practice still is likely to be
quite low.
Potential for Harm
We were unable to find literature that implicated checklists as causes of adverse events.
There exists a theoretical possibility of delays due to complex checklists, but these have not been
borne out in any published studies. There has also been concern raised about compulsory
checklist procedures as an unnecessary duplication of work already performed by operating
room technical personnel.
Costs and Implementation
The FDA checklist, or local variants thereof, is widely implemented and inexpensive.
Although several authors have mentioned that checkout processes are not used in 100% of cases,
this does not seem to reflect problems with checklists themselves. 4,12,13 The work of March and
Crowley 3 suggests that checklists training may be critically important. However, the most
important barrier to implementation is the heterogeneity of anesthesia delivery devices
themselves, which makes creation of a single, effective, broadly generalizable checklist difficult
if not impossible.
Comment
Given its face validity and the theoretical connection between anesthesia checklists and
those used so effectively in aviation, preanesthesia checklists represent plausible safety tools.
However, the reliability of modern anesthesia machines has reduced the frequency of mechanical
failure to such a degree that adverse outcomes due to anesthesia machine failure are exceedingly
rare. The checklists examined in the cited studies only examine the integrity of the anesthesia
machine and ventilatory monitors, not cardiovascular monitors, airway equipment, intravenous
apparatus, infusion pumps, or medications. Standardized checklists for these other critical
components of anesthesia care do not exist in the literature. This may explain the paucity of
literature exploring the use of checklists and their real effects on patient outcomes. Furthermore,
the inability of anesthesiologists to detect preset faults in the cited studies may simply reflect the
infrequency with which such faults are encountered in modern anesthesia practice.
The face validity of checklists and the difficulty of “probing” their value in clinical
studies make additional “proof of concept” studies unlikely. Although future investigations could
not ethically study the anesthesia machine checklist per se, they could seek to determine more
effective methods for its implementation, or could develop additional checklists with a broader
scope. The little evidence we have been able to uncover suggests that, like the use of voluntary
guidelines elsewhere in medicine (Chapter 51), checklists are not used uniformly. This may
result from a sense that these checks are low-yield, redundant, onerous, or all of the above.
The need for effective checkout procedures is likely to grow as the complexity of
anesthesia equipment increases. This will increase the need to make checklists more sensitive
and specific in detecting faults, while improving usability. These worthy goals may then serve as
templates for other technologically dependent medical specialties, such as cardiac
electrophysiology, and interventional radiology.
261

Table 23.1. Evaluations of the FDA “checkout list” for preoperative anesthesia equipment
assessment
Study Study Design,
Outcomes
March, 1991 3 : Exposure of a total of
188 anesthesiology residents and
practitioners from multiple sites to a
“mobile anesthesia study center” with 2
anesthesia machines pre-set to one of
two different “fault sets.” Each of these
sets included 4 independent machine
faults. Practitioners were instructed to
use their own checkout methods to
assess a machine with one of the “fault
sets.” After completion of this
assessment, the machine was adjusted to
display the other fault set, and the
participant invited to use the FDA
checkout list in the second assessment.
Manley, 1996 9 : Similar study to above,
but involving only 22 participants (a
mixture of anesthesiologist, nurse
anesthetists, and senior nurse anesthetist
students) from only one site.
Mixture of
Levels 2 & 3,
Level 2
262
Results
Participants detected an average of
only 1 in 4 machine faults. A
statistically significant
improvement in the fault detection
rate with use of the FDA checkout
compared with the participants’
individual methods was observed
for only 1 of the 4 fault types
(involving the oxygen/nitrous
oxide ratio).
As above For both of the fault sets,
approximately half of the
participants detected fewer than
50% of the 4 target faults using
their own individual methods. The
detection rates using the revised
FDA checkout list did not differ
significantly (p=0.48) from these
results.

References
1. Reason JT. Human Error. New York: Cambridge Univ Press; 1990.
2. Cooper JB, Newbower RS, Kitz RJ. An analysis of major errors and equipment failures in
anesthesia management: considerations for prevention and detection. Anesthesiology.
1984;60:34-42.
3. March MG, Crowley JJ. An evaluation of anesthesiologists' present checkout methods and
the validity of the FDA checklist. Anesthesiology. 1991;75:724-729.
4. Charlton JE. Checklists and patient safety. Anaesthesia. 1990;45:425-426.
5. Walker JS, Wilson M. Clinical risk management in anaesthesia. Qual Health Care.
1995;4:115-121.
6. Anesthesia Apparatus Checkout Recommendations, 1993. Fed Regist. 1994;59:35373-74.
7. U.S. Food and Drug Adminstration. Anesthesia apparatus checkout recommendations.
Available at: http://www.fda.gov/cdrh/humfac/anesckot.pdf. Accessed June 10, 2001.
8. Cooper JB, Newbower RS, Long CD, McPeek B. Preventable anesthesia mishaps: a study
of human factors. Anesthesiology. 1978;49:399-406.
9. Spooner RB, Kirby RR. Equipment-related anesthetic incidents. Int Anesthesiol Clin.
1984;22:133-147.
10. Manley R, Cuddeford JD. An assessment of the effectiveness of the revised FDA checklist.
AANA J. 1996;64:277-282.
11. Degani A, Wiener EL. Cockpit checklists - concepts, design, and use. Human Factors.
1993;35:345-359.
12. Cooper JB, Cullen DJ, Eichhorn JH, Philip JH, Holzman RS. Administrative guidelines for
response to an adverse anesthesia event. The Risk Management Committee of the Harvard
Medical School's Department of Anaesthesia. J Clin Anesth. 1993;5:79-84.
13. Currie M. A prospective survey of anaesthetic critical events in a teaching hospital.
Anaesth Intensive Care. 1989;17:403-411.
263

264

Chapter 24. The Impact Of Intraoperative Monitoring On Patient Safety
Salim D. Islam, MD
Andrew D. Auerbach, MD, MPH
University of California, San Francisco School of Medicine
Background
Until the 1960s, intraoperative monitoring consisted of a blood pressure cuff,
electrocardiogram (ECG), stethoscope, and the vigilance of the anesthesiologist. Over the next 2
decades, the array of available monitors burgeoned, and clinical practice varied widely. In 1986,
in an effort to improve patient safety, standards for intraoperative monitoring were developed
and implemented by the American Society of Anesthesiologists (ASA). 1 They have been almost
universally adopted by anesthesia providers in the United States and now form the standard of
care in this country. The ASA standards are summarized in Figure 24.1.
Concurrently with the implementation of better monitoring, anesthesia-related mortality
has fallen sharply. Proponents of monitoring claim that better monitoring is the reason for
improvement in patient safety. 2-4 Others have claimed that advances in knowledge and training
combined with the development of safer medications have had as much impact on patient safety
as the adoption of monitoring standards. 5,6 In this chapter, we evaluate the evidence linking
intraoperative monitoring to patient safety.
Practice Description
Intraoperative monitoring involves the use of mechanical devices to record and display
physiologic parameters such as heart rate, blood pressure, oxygen saturation, and temperature.
Standard routine monitoring is noninvasive, employing blood pressure cuff, ECG, and pulse
oximetry.
Invasive monitors such as arterial and central venous catheters and transesophageal
echocardiography may provide more detailed and timely physiologic information, but also pose
an increased risk for iatrogenic complications. In practice these monitors are used selectively,
and are not reviewed here.
Prevalence and Severity of the Target Safety Problem
Death due to anesthesia has become rare. In one large Canadian study involving 27,184
inpatients who underwent anesthesia, physician review of 115 randomly selected “major events”
classified less than 20% as having any anesthetic involvement, with no deaths even partially
attributed to anesthesia. 7 In the United States, the mortality due to general anesthesia has been
estimated at approximately 5000 deaths per year (in the 1970s), 8 with approximately half that
number estimated in the 1980s. 9 Thus, general anesthesia represents the one aspect of health care
where the risk of death is low enough to rival the safety record achieved in other high-risk
industries such as aviation. 10
By contrast, morbid events (complications) related to anesthetic care are likely more
prevalent and difficult to classify as preventable or unavoidable. Because certain aspects of
monitoring may reduce the incidence of morbid events unrelated to anesthesia, assessing the
impact of monitoring practices solely on anesthetic outcomes may be inappropriate. For
example, detection of a consistent decrease in intraoperative blood pressure may signal
unrecognized bleeding, allowing the anesthesiologist to alert the surgeon to this possibility and
prompting appropriate management. While intraoperative hemorrhage does not represent an
265

anesthetic complication, intraoperative blood pressure monitoring can clearly contribute to the
overall safety of the surgical patient. Thus, the scope of intraoperative morbidity targeted by
anesthetic monitoring practices is much broader than the set of possible complications
attributable solely to the administration of anesthesia. 7-9
Opportunities for Impact
In the United States, there are no mandatory regulations for monitoring practices.
However, virtually all anesthesiologists abide by the monitoring standards set forth by the 1986
ASA standards, last modified in 1998 (Figure 24.1). Although these standards were implemented
with only speculative evidence of their benefit, 4 few clinicians doubt their merit.
Study Designs
Using a structured MEDLINE search, we identified articles presenting data related to the
impact of perioperative monitoring. Many of these studies 11-17 involved Level 4 designs (eg,
observational studies without a control group). For instance, several of the articles 11-13,15 reported
data from the Australian Incident Monitoring Study and involved analysis of a case series of
2000 incident reports without accompanying controls. Other studies only indirectly pertained to
intraoperative monitoring. One study surveyed anesthesiologists regarding their views on the
appropriate alarm settings for intraoperative blood pressure monitoring. 18 Another focused on the
personnel performing intraoperative monitoring—physician anesthesiologists versus certified
nurse anesthetists. 19 We chose not to purse this contentious and intensely political comparison,
as few studies have compared the outcomes achieved by these two groups. Moreover, our
reviewer team did not include a nurse anesthetist, making any conclusions drawn more
susceptible to bias. Of the 3 remaining studies, one involved a non-randomized clinical trial
(Level 2), but a Level 3 outcome. 20
The remaining 2 studies met our inclusion criteria (Chapter 3). One was a retrospective
analysis of anesthesia accidents before and after the implementation of monitoring standards
(Level 3), 2 and the other used a prospective, randomized, controlled trial design (Level 1) to
assess the impact of pulse oximetry on postoperative complications (Level 1 outcome). 21
Study Outcomes
The 2 studies 2,21 meeting the methodologic inclusion criterion reported morbidity and
mortality (Level 1) attributable to anesthesia, ie, a major complication or death occurring in the
immediate postoperative period not obviously explained by the patient’s underlying condition or
the operation itself.
Evidence for Effectiveness of the Practice
Through a review of cases reported to a liability insurer, Eichhorn identified 11 major
intraoperative accidents solely attributable to anesthesia among over 1,000,000 cases performed
at the nine Harvard hospitals from 1976-1988. 2 Eight of these accidents were judged to be
preventable as they were caused by failure to ventilate or to deliver adequate oxygen to the
patient. Only one of these accidents occurred after the adoption of monitoring standards in mid-
1985, supporting the safety benefit of intraoperative monitoring standards, although the
difference in accident frequency was not statistically significant.
In a multicenter, randomized, controlled trial of 20,802 surgical patients, Moller et al 21
studied the impact of perioperative pulse oximetry on patient outcome. Despite the large sample,
the authors were unable to show a difference in in-hospital mortality or postoperative
266

complications. During anesthesia and in the post-anesthesia care unit (PACU), more episodes of
hypoxemia and myocardial ischemia were detected in patients monitored with pulse oximetry. 21
Potential for Harm
Routine noninvasive monitoring carries minimal (although not zero) additional risk for
iatrogenic complications from the devices themselves. Current standard of practice requires that
they be used in all cases of general or regional anesthesia. However the number of monitors and
their concomitant alarms raises the possibility of additional harm. A study of monitor alarms in
the intensive care unit (ICU) suggested that monitor alarms might actually reduce quality of care
because of their high frequency and low specificity. In this study, an alarm occurred every 37
minutes, and in the majority of cases (72%) no change in management was indicated as a
result. 22
Costs and Implementation
The costs of intraoperative monitors are largely fixed in the acquisition cost of the
monitoring device. Incremental patient costs for disposables are minimal.
Comment
The inability of a very large multicenter study 21 to detect a benefit in morbidity and
mortality from pulse oximetry—by all accounts the most useful monitor—suggests that the
magnitude of benefit may be so small that an adequate study to detect this difference may not be
feasible. Along with capnography (carbon dioxide monitoring), pulse oximetry is often cited as
the monitoring method most able to detect potential critical incidents early enough to prevent
adverse outcomes. 2,6 This conjecture is supported by the ASA Closed Claims Study. Analyzing
1175 claims, the study concluded that the combination of pulse oximetry and capnography
“could be expected” to help prevent anesthetic-related morbidity and mortality. 23
Despite a lack of randomized trial data, the practice of noninvasive intraoperative
monitoring has become standard of care. This has resulted from the ASA Monitoring Standards
and physicians' faith in the practice based on its face value, along with some confirmatory
evidence drawn from incident reporting systems. 11,16,17 As such, it seems likely that future
research into intraoperative monitoring will be unable to study approaches that do not include
standard, noninvasive monitoring. Future investigation might seek to determine which
monitoring methods detect “near misses” more effectively.
Moving beyond non-invasive techniques, there is a great need to identify which
specialized monitors provide a safety benefit in selected patient populations. The use of
pulmonary artery catheters for monitoring critically ill patients represents a well-known example
of a practice with substantial face validity but unclear impact on patient outcomes. 24,25 In
addition, new, noninvasive alternatives to invasive monitors (eg, esophageal or spirometry-based
cardiac output monitors) may ultimately allow us to obtain the same information at less risk to
the patient.
267

Figure 24.1. ASA standards for basic anesthetic monitoring*
Standard 1: Qualified anesthesia personnel shall be present in the room
throughout the conduct of all general anesthetics, regional anesthetics and
monitored anesthesia care
Standard 2: During all anesthetics, the patient’s oxygenation, ventilation,
circulation, and temperature shall be continually* evaluated
Oxygenation:
Oxygen analyzer for inspired gases
Observation of the patient
Pulse oximetry
Ventilation:
Auscultation of breath sounds
Observation of the patient
Observation of the reservoir bag
Capnography (Carbon dioxide monitoring)
Circulation:
Continuous* ECG display
Heart rate and BP recorded every 5 minutes
Evaluation of circulation
Auscultation of heart sounds
Palpation of pulse
Pulse plethysmography
Pulse oximetry
Intraarterial pressure tracing
Temperature:
Monitor temperature when changes are intended, anticipated, or suspected
*The term “continuous” means prolonged without interruption; "continually"
means repeated regularly and frequently. ECG indicates electrocardiography;
BP, blood pressure.
268

References
1. American Society of Anesthesiologists. Standards of the American Society of
Anesthesiologists: Standards for Basic Anesthetic Monitoring. Available at:
http://www.asahq.org/Standards/02.html. Accessed May 23, 2001.
2. Eichhorn JH. Prevention of intraoperative anesthesia accidents and related severe injury
through safety monitoring. Anesthesiology. 1989;70:572-577.
3. Eichhorn JH. Effect of monitoring standards on anesthesia outcome. Int Anesthesiol Clin.
1993;31:181-196.
4. Keenan RL, Boyan CP. Cardiac arrest due to anesthesia. A study of incidence and causes.
JAMA. 1985;253:2373-2377.
5. Orkin FK. Practice standards: the Midas touch or the emperor's new clothes?
Anesthesiology. 1989;70:567-571.
6. Brodsky JB. What intraoperative monitoring makes sense? Chest. 1999;115:101S-105S.
7. Cohen MM, Duncan PG, Pope WD, Biehl D, Tweed WA, MacWilliam L, et al. The
Canadian four-centre study of anaesthetic outcomes: II. Can outcomes be used to assess the
quality of anaesthesia care? Can J Anaesth. 1992;39:430-439.
8. Phillips OC, Capizzi LS. Anesthesia mortality. Clin Anesth. 1974;10:220-244.
9. Deaths during general anesthesia: technology-related, due to human error, or unavoidable?
An ECRI technology assessment. J Health Care Technol. 1985;1:155-175.
10. National Transportation Safety Board. Aviation Accident Statistics. Available at:
http://www.ntsb.gov/aviation/Stats.htm. Accessed May 28, 2001.
11. Webb RK, van der Walt JH, Runciman WB, Williamson JA, Cockings J, Russell WJ, et al.
The Australian Incident Monitoring Study. Which monitor? An analysis of 2000 incident
reports. Anaesth Intensive Care. 1993;21:529-542.
12. Klepper ID, Webb RK, Van der Walt JH, Ludbrook GL, Cockings J. The Australian
Incident Monitoring Study. The stethoscope: applications and limitations–an analysis of
2000 incident reports. Anaesth Intensive Care. 1993;21:575-578.
13. Cockings JG, Webb RK, Klepper ID, Currie M, Morgan C. The Australian Incident
Monitoring Study. Blood pressure monitoring–applications and limitations: an analysis of
2000 incident reports. Anaesth Intensive Care. 1993;21:565-569.
14. Hewer I, Drew B, Karp K, Stotts N. The utilization of automated ST segment analysis in
the determination of myocardial ischemia. AANA J. 1997;65:351-356.
15. Williamson JA, Webb RK, Cockings J, Morgan C. The Australian Incident Monitoring
Study. The capnograph: applications and limitations–an analysis of 2000 incident reports.
Anaesth Intensive Care. 1993;21:551-557.
16. Spittal MJ, Findlay GP, Spencer I. A prospective analysis of critical incidents attributable
to anaesthesia. Int J Qual Health Care. 1995;7:363-371.
17. Findlay GP, Spittal MJ, Radcliffe JJ. The recognition of critical incidents: quantification of
monitor effectiveness. Anaesthesia. 1998;53:595-598.
18. Asbury AJ, Rolly G. Theatre monitor alarm settings: a pilot survey in Scotland and
Belgium. Anaesthesia. 1999;54:176-180.
19. Silber JH, Kennedy SK, Even-Shoshan O, Chen W, Koziol LF, Showan AM, et al.
Anesthesiologist direction and patient outcomes. Anesthesiology. 2000;93:152-163.
20. Kay J, Neal M. Effect of automatic blood pressure devices on vigilance of anesthesia
residents. J Clin Monit. 1986;2:148-150.
269

21. Moller JT, Johannessen NW, Espersen K, Ravlo O, Pedersen BD, Jensen PF, et al.
Randomized evaluation of pulse oximetry in 20,802 patients: II. Perioperative events and
postoperative complications. Anesthesiology. 1993;78:445-453.
22. Chambrin MC, Ravaux P, Calvelo-Aros D, Jaborska A, Chopin C, Boniface B.
Multicentric study of monitoring alarms in the adult intensive care unit (ICU): a descriptive
analysis. Intensive Care Med. 1999;25:1360-1366.
23. Tinker JH, Dull DL, Caplan RA, Ward RJ, Cheney FW. Role of monitoring devices in
prevention of anesthetic mishaps: a closed claims analysis. Anesthesiology. 1989;71:541-
546.
24. Connors AF, Speroff T, Dawson NV, Thomas C, Harrell FE, Wagner D, et al. The
effectiveness of right heart catheterization in the initial care of critically ill patients.
SUPPORT Investigators. JAMA. 1996;276:889-897.
25. Ivanov R, Allen J, Calvin JE. The incidence of major morbidity in critically ill patients
managed with pulmonary artery catheters: a meta-analysis. Crit Care Med. 2000;28:615-
619.
270

Chapter 25. Beta-blockers and Reduction of Perioperative Cardiac Events
Andrew D. Auerbach MD, MPH
University of California, San Francisco School of Medicine
Background
As the most common complications of major noncardiac surgery, myocardial infarction
and cardiovascular death have long been a focus of preoperative evaluations 1-4 and a target of
perioperative management strategies. Until recently, methods to reduce the incidence of these
complications depended upon preoperative assessments of risk combining clinical evaluation
with clinical prediction rules, followed by additional tests or revascularization procedures, as
appropriate. 1 The benefit of preoperative revascularization remains unclear, as no randomized
prospective trial has demonstrated its benefit. 5 Indeed, concern exists that preoperative
intervention might prove detrimental, as the net benefit in terms of reduced perioperative cardiac
events may be offset by the risks of the revascularization strategy itself. Newer strategies,
including the use of percutaneous transluminal angioplasty as the revascularization modality,
may have promise. 6 Large prospective trials examining these approaches are underway. 5
Strong evidence links myocardial ischemia with postoperative myocardial events. 7,8 One
study found postoperative ischemia increased the odds of postoperative myocardial events 21fold.
9 Based on findings from observational studies that beta-blockade blunts
electrocardiographic signs of ischemia, 10-12 recent trials have examined the effects of
perioperative beta-blocker administration on patient outcomes. Results of these investigations
are extremely promising, and beta-blockade may represent an important new method of reducing
perioperative cardiac risk. This chapter reviews the evidence from randomized controlled trials
examining the effect of perioperative beta-blockade on cardiac events (ie, myocardial ischemia,
angina, myocardial infarction, pulmonary edema, and cardiac death).
Practice Description
Although published studies have employed different agents, doses, and dosing schedules,
the general approach in each study has been similar: administration of a therapeutic dose of betablocker
prior to induction of anesthesia, followed by beta-blockade through the operation and in
the postoperative period. In all regimens, the dose is titrated to a target heart rate, generally 70
beats per minute or lower.
Prevalence and Severity of the Target Safety Problem
Myocardial cardiac events are the most common medical complication of surgery,
occurring in 2-5% of patients undergoing non-cardiac surgery 13 and as many as 30% of patients
undergoing vascular surgery. 14,15 Perioperative cardiac events are associated with a mortality rate
of nearly 60% per event, 14,16 prolonged hospitalization, and higher costs. 16,17 The prevalence of
these events and their high mortality have made the prevention of perioperative cardiac ischemia
the subject of practice guidelines 1,17 and numerous prediction rules 13,16,18 to detect patients at
high risk for these complications.
271

Opportunities for Impact
As a relatively new therapy, few data describe the use of perioperative beta-blockade in
clinical practice. However, evidence suggests it is utilized infrequently. A recent observational
study in the Netherlands of patients undergoing vascular surgery showed that only 27% of these
high-risk patients received beta-blockers perioperatively. 19
Study Designs
Using a structured MEDLINE search, we identified 4 relevant randomized controlled
trials of the effectiveness of perioperative beta-blockade in reducing perioperative cardiac
events, including myocardial ischemia and cardiac or all-cause mortality (Table 25.1). A
randomized trial by Harwood et al was excluded because both groups received beta-blockers (ie,
there was no control group). 20 Although data from a study by Wallace et al 21 were derived from
one of the randomized trials included in this review, 22 it reported effects of beta-blockade upon
different outcomes (ie, myocardial ischemia) and was included in our review. There was
sufficient evidence available to limit the review to studies of Level 1 design. Observational
studies, such as those by Pasternack et al and Boersma et al, 11,19 are not included.
Study Outcomes
The studies identified included a range of clinical outcomes: 2 included assessment of
myocardial ischemia 12,23 and 3 reported myocardial infarction, pulmonary edema, cardiac death,
or all-cause mortality (Level 1 outcomes). 15,22,23
Evidence for Effectiveness of the Practice
Of studies reporting the effect of beta-blockers on perioperative ischemia (Level 2
outcome), all but one found a statistically significant reduction in ischemia among treated
patients. Wallace et al, 21 in a subset analysis of data from Mangano et al, 24 reported less frequent
perioperative myocardial ischemia in atenolol-treated patients. Stone et al 25 suggested a similar
effect of beta-blockade on Holter-monitor documented myocardial ischemia. However, the
authors did not report the types of procedures included in their sample, nor did they statistically
compare baseline patient characteristics, leaving their conclusions open to debate. Raby et al 12
also found a significant beneficial effect of beta-blockade using a continuous infusion of esmolol
in high-risk patients undergoing vascular surgery. Although Urban et al also found a reduction in
perioperative ischemia, this difference failed to reach statistical significance. 23 These findings
may be explained in part by a relatively low cardiac risk in Urban’s cohort, who were
undergoing elective total knee replacement. The patients in many of the other studies were at
higher risk of cardiac events, as demonstrated by rates of ischemia in the control groups. In
studies finding a statistical difference, rates of ischemia were between 28% and 73% in controls,
as compared with the 15% rate of ischemia observed in Urban's control group.
Of studies reporting cardiac events and cardiac mortality, 2 reported significant
improvement in patient outcomes due to beta-blockade. In a study of male veterans undergoing
major noncardiac surgery, Mangano et al 22 reported a relative reduction in all-cause mortality of
nearly 55% at 2 years. This difference, which appeared within the first 8 months of follow-up,
was ascribed to a marked reduction in cardiac events in the first year of therapy (67% reduction
at year 1, 48% at year 2). However, patients in the beta-blocker group had less coronary disease
at study entry, were on ACE-inhibitors more frequently, and were less likely to have betablockers
discontinued perioperatively, perhaps biasing results in favor of the treatment group. 26,
272

27
Accounting for these differences in multivariate models of varying stringency did not
invalidate their findings. 24 Although questions remain about the generalizability of results to
other patient populations, the authors favored broader use of beta-blockade in the setting of
clinical trials.
Poldermans et al 15 suggested an even greater benefit of beta-blockade among high-risk
patients. These investigators enrolled patients undergoing vascular surgery who had myocardial
ischemia documented by dobutamine echocardiography, with an estimated rate of perioperative
cardiac event of 28%. The entire patient cohort experienced a 90% reduction in cardiac death or
non-fatal myocardial infarction at 30 days. Follow-up care did not include additional therapy (ie,
cardiac catheterization, revascularization), raising concerns that the research algorithm did not
reflect optimal clinical practice. 28, 29 However, if the true rate of events in treated patients is low
(the point estimate from this small study was 3.4%), the risks associated with revascularization 30
may outweigh any benefit.
In contrast to the previous 2 studies, Urban et al 23 found no statistically significant
difference in rates of in-hospital myocardial infarction. It is likely that these investigators' ability
to detect a difference was limited in part by the relatively small sample size and shorter length of
follow-up. Other studies of perioperative beta-blockade employed longer periods of follow-up to
detect events up to 2 years following surgery.
Differences in absolute magnitude of benefit can be ascribed in part to the cardiac risks
of the patients enrolled (again reflected in event rates in the control groups in each study), with
the most powerful benefits seen in patients at highest risk. The greater benefit seen in
Poldermans et al’s study 15 may also be due to the fact that the study did not enroll patients who
were receiving beta-blockers. Patients who are beta-blocker naïve may have a different response
to perioperative use of bisoprolol, or the preexisting use of these agents may represent a
confounding factor not completely accounted for in other studies of perioperative beta-blockade.
Beta-blockade may have additional beneficial effects for elderly patients. Patients who
received beta-blockers were extubated more quickly, required less medication for pain, and were
alert sooner after surgery. 31 Although the unblinded nature of this study leaves its findings open
to debate, the possibility of additional benefits is tantalizing and worthy of further investigation.
Potential for Harm
Stone et al reported high rates of bradycardia (21/89 patients) in beta-blocker treated
patients, “half” of whom required atropine therapy. However, the vague descriptions and more
general problems with the study's design make it difficult to interpret the significance of these
events in clinical practice. Adverse events related to the use of beta-blockers in other reviewed
studies were infrequent (10% or less in Mangano et al 22 ) and did not require therapy or result in
withdrawal of the medication. Similar rates of side effects have been noted in studies examining
beta-blockade in patients undergoing cardiac surgery. 20,32,33 One study examining use of
propranolol in patients undergoing thoracotomy for pneumonectomy suggested that patients
receiving beta-blockers had twice the rate of postoperative congestive heart failure (4/50 vs.
8/50, p

Costs and Implementation
The costs of beta-blockers are generally low, and the systems required to use them
according to the protocols used in these studies are already in place. In addition, there is the
potential for significant cost-savings if routine use of beta-blockers allows a safe reduction in the
use of extensive preoperative cardiovascular testing.
Comment
Results from several well-designed clinical trials suggest that use of beta-blockers in the
perioperative period is associated with significant reductions in patient cardiac morbidity and
mortality. In the future such therapy may reduce the need for additional tests and
revascularization procedures, 14 further reducing costs of care. However, several questions
regarding its use remain, and should be topics of future research.
First, no clear data suggest an advantage of one particular beta-blocking agent over
another. Studies to date have employed several different beta-blockers, suggesting that the
efficacy of beta-blockade is class dependent if titrated to physiologically active dosages. Other
(alpha-1 selective) sympatholytics also improve patient outcomes, 36 raising the possibility that
combined alpha-beta antagonists (ie, labetolol) may have benefit. Second, results from
Shammash et al document the hazards of discontinuation of beta-blockers immediately
postoperatively, 35 and most protocols employed treatment regimens that extended longer - even
up to one month following surgery. The current studies suggest beta-blockade should be
continued for at least one week postoperatively. Third, effectiveness of beta-blockade in patients
at high risk due to aortic stenosis or unstable or severe cardiovascular symptoms (New York
Heart Association Class III-IV) is unknown, as these patients were not included in the reviewed
studies. Similarly, its utility - both in terms of cardiac events and cost - in patients with very low
risk of perioperative cardiac events (ie, those undergoing same-day or outpatient surgery,
ophthalmic surgery, or those who have minimal cardiac risk) is unclear. Beta-blockade has not
been studied in patients undergoing regional anesthesia or conscious sedation. In addition, no
study to date has examined the use of beta-blockade in patients who have poor functional status
and might otherwise be referred for additional non-invasive testing. 1,5,14,17
Finally, the increasing popularity of perioperative beta-blockade, particularly catalyzed
by the results of the study by Poldermans et al, 15 calls into question whether risk stratification
using published guidelines or risk indices is still necessary. 28 Although beta-blockade is likely to
be effective in many patients, the identification of patients at highest risk is still important, as
these patients may require additional testing and therapy. A recent study of beta-blockade noted
improved outcomes across a spectrum of predicted cardiac risk, but noted that cardiac events
could be further reduced in high-risk patients through use of additional non-invasive testing and
subsequent “usual care.” 19 Thus, although beta-blockade may increase the threshold at which
clinicians refer patients for additional testing, the era of risk stratification is not over.
274

The use of beta-blockers to reduce perioperative cardiac events and mortality represents a
major advance in perioperative medicine for some patients at intermediate and high risk for
cardiac events during noncardiac surgery. Wider use of this therapy should be promoted and
studied, with future research focused on fine-tuning dosages and schedules and identifying
populations of patients in which its use is cost-effective.
275

Table 25.1. Randomized controlled trials of the effectiveness of perioperative beta-blockade*
Study Participants Regimen Results† Side Effects Comments
Mangano,
1996 22
Wallace,
1998 21
Polderman
s, 1999 15
Raby,
1999 12
Stone,
1988 37
200 patients
undergoing
elective
noncardiac
surgery
112 patients with
positive results on
dobutamine
echocardiography
undergoing
elective
abdominal aortic
or infrainguinal
arterial
reconstruction
26 patients with
preoperative
ischemia by
Holter monitor
undergoing aortic
aneurysm repair,
infrainguinal
arterial bypass, or
carotid
endarterectomy
128 untreated
hypertensive
patients
undergoing
elective surgery.
Hypertension
Atenolol 5-10g IV
30 min before entry
into OR, after
surgery, and 50-
100g qd through
hospital stay (up to
7 days); Target HR
55-65 bpm; doses
held if HR

Urban,
2000 23
defined as systolic
blood pressure
160-200 mmHg,
diastolic 90-100
mmHg
120 patients
undergoing
elective total knee
arthroplasty
mg po given before
induction of
anesthesia
Esmolol IV within
1 hr after surgery,
titrate to HR

10. Smulyan H, Weinberg SE, Howanitz PJ. Continuous propranolol infusion following
abdominal surgery. JAMA. 1982;247:2539-2542.
11. Pasternack PF, Grossi EA, Baumann FG, et al. Beta blockade to decrease silent myocardial
ischemia during peripheral vascular surgery. Am J Surg. 1989;158:113-116.
12. Raby KE, Brull SJ, Timimi F, et al. The effect of heart rate control on myocardial ischemia
among high-risk patients after vascular surgery. Anesth Analg. 1999;88:477-482.
13. Lee TH, Marcantonio ER, Mangione CM, et al. Derivation and prospective validation of a
simple index for prediction of cardiac risk of major noncardiac surgery. Circulation.
1999;100:1043-1049.
14. Lee TH. Reducing cardiac risk in noncardiac surgery. N Engl J Med. 1999;341:1838-1840.
15. Poldermans D, Boersma E, Bax JJ, et al. The effect of bisoprolol on perioperative mortality
and myocardial infarction in high-risk patients undergoing vascular surgery. Dutch
Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography Study
Group. N Engl J Med. 1999;341:1789-1794.
16. Goldman L. Multifactorial index of cardiac risk in noncardiac surgery: ten-year status
report. J Cardiothorac Anesth. 1987;1:237-244.
17. Eagle KA, Brundage BH, Chaitman BR, et al. Guidelines for perioperative cardiovascular
evaluation for noncardiac surgery: an abridged version of the report of the American
College of Cardiology/American Heart Association Task Force on Practice Guidelines.
Mayo Clin Proc.1997;72:524-531.
18. Detsky AS, Abrams HB, Forbath N, Scott JG, Hilliard JR. Cardiac assessment for patients
undergoing noncardiac surgery. A multifactorial clinical risk index. Arch Intern Med.
1986;146:2131-2134.
19. Boersma E, Poldermans D, Bax JJ, et al. Predictors of cardiac events after major vascular
surgery: Role of clinical characteristics, dobutamine echocardiography, and beta-blocker
therapy. JAMA. 2001;285:1865-1873.
20. Harwood TN, Butterworth J, Prielipp RC, et al. The safety and effectiveness of esmolol in
the perioperative period in patients undergoing abdominal aortic surgery. J Cardiothorac
Vasc Anesth. 1999;13:555-561.
21. Wallace A, Layug B, Tateo I, et al. Prophylactic atenolol reduces postoperative myocardial
ischemia. McSPI Research Group. Anesthesiology. 1998;88:7-17.
22. Mangano DT, Layug EL, Wallace A, Tateo I. Effect of atenolol on mortality and
cardiovascular morbidity after noncardiac surgery. Multicenter Study of Perioperative
Ischemia Research Group. N Engl J Med. 1996;335:1713-1720.
23. Urban MK, Markowitz SM, Gordon MA, Urquhart BL, Kligfield P. Postoperative
prophylactic administration of beta-adrenergic blockers in patients at risk for myocardial
ischemia. Anesth Analg. 2000;90:1257-1261.
24. Mangano DT. Effect of atenolol on mortality and cardiovascular morbidity after noncardiac
surgery. N Engl J Med. 1997;336:1452.
25. Stone JG, Foex P, Sear JW, Johnson LL, Khambatta HJ, Triner L. Myocardial ischemia in
untreated hypertensive patients: effect of a single small oral dose of a beta-adrenergic
blocking agent. Anesthesiology. 1988;68:495-500.
26. Reis SE, Feldman AH. Effect of atenolol on mortality and cardiovascular morbidity after
noncardiac surgery. N Engl J Med. 1997;336:1453.
27. Petros JA. Effect of atenolol on mortality and cardiovascular morbidity after noncardiac
surgery. N Engl J Med. 1997;336:1452.
278

28. Litwack R, Gilligan D, DeGruttola V. Beta-Blockade for patients undergoing vascular
surgery. N Engl J Med. 2000;342:1051-1053.
29. Feldman T, Fusman B, McKinsey JF. Beta-Blockade for patients undergoing vascular
surgery. N Engl J Med. 2000;342:1051-1052.
30. Poldermans D, Boersma E. Beta-Blockade for patients undergoing vascular surgery. N
Engl J Med. 2000;342:1052-1053.
31. Zaugg M, Tagliente T, Lucchinetti E, et al. Beneficial effects from beta-adrenergic
blockade in elderly patients undergoing noncardiac surgery. Anesthesiology. 1999;91:1674-
1686.
32. Hammon JW, Wood AJ, Prager RL, Wood M, Muirhead J, Bender HW. Perioperative beta
blockade with propranolol: reduction in myocardial oxygen demands and incidence of
atrial and ventricular arrhythmias. Ann Thorac Surg. 1984;38:363-367.
33. Lamb RK, Prabhakar G, Thorpe JA, Smith S, Norton R, Dyde JA. The use of atenolol in
the prevention of supraventricular arrhythmias following coronary artery surgery. Eur
Heart J. 1988;9:32-36.
34. Bayliff CD, Massel DR, Inculet RI, et al. Propranolol for the prevention of postoperative
arrhythmias in general thoracic surgery. Ann Thorac Surg. 1999;67:182-186.
35. Shammash JB, Trost JC, Gold JM, Berlin JA, Golden MA, Kimmel SE. Perioperative betablocker
withdrawal and mortality in vascular surgical patients. Am Heart J. 2001;141:148-
153.
36. Oliver MF, Goldman L, Julian DG, Holme I. Effect of mivazerol on perioperative cardiac
complications during non-cardiac surgery in patients with coronary heart disease: the
European Mivazerol Trial (EMIT). Anesthesiology. 1999;91:951-961.
37. Stone JG, Foex P, Sear JW, Johnson LL, Khambatta HJ, Triner L. Risk of myocardial
ischaemia during anaesthesia in treated and untreated hypertensive patients. Br J Anaesth.
1988;61:675-679.
279

Section D. Safety Practices for Hospitalized or Institutionalized Elders
Chapter 26. Prevention of Falls in Hospitalized and Institutionalized Older People
Subchapter 26.1. Identification Bracelets for High-Risk Patients
Subchapter 26.2. Interventions that Decrease the Use of Physical Restraints
Subchapter 26.3. Bed Alarms
Subchapter 26.4. Special Hospital Flooring Materials to Reduce Injuries from
Patient Falls
Subchapter 26.5. Hip Protectors to Prevent Hip Fracture
Chapter 27. Prevention of Pressure Ulcers in Older Patients
Chapter 28. Prevention of Delirium in Older Hospitalized Patients
Chapter 29. Multidisciplinary Geriatric Consultation Services
Chapter 30. Geriatric Evaluation and Management Units for Hospitalized Patients
280

Chapter 26. Prevention of Falls in Hospitalized and Institutionalized
Older People
Joseph V. Agostini, MD
Dorothy I. Baker, PhD, RNCS
Sidney T. Bogardus, Jr., MD
Yale University Schools of Medicine and Public Health
Introduction
A fall is defined as unintentionally coming to rest on the ground, floor, or other lower
level, but not as a result of syncope or overwhelming external force. Falling is a common cause
of morbidity and the leading cause of nonfatal injuries and trauma-related hospitalizations in the
United States. 1 Complications include bone fractures, injury to the soft tissues, increased
functional dependence, and fear of falling again, which itself can be debilitating. Each of these
complications contributes to increased risk of future falls. Studies in community-dwelling older
patients have identified age, gait or balance impairment, sensory or cognitive impairment,
musculoskeletal diseases, environmental hazards, and many medications (such as sedativehypnotic
drugs) as risk factors.
One of the strongest predictors of future falls is having previously fallen. There are
numerous other risk factors for falls in older persons, which are reviewed in detail elsewhere. 2,3
The number of risk factors is correlated with the risk of falling. A study by Tinetti and
colleagues found the risk of falling increased from 19% when one risk factor was present to 78%
in the presence of 4 or more risk factors. 4 Some of the factors associated with fall risk in the
hospital setting, however, may differ from those in community-dwelling or institutional settings.
The onset of an acute illness leading to hospitalization may increase fall risk due to immobility
and deconditioning. Treatment for an acute condition, such as the addition of new medications or
an altered medication regimen, may also increase fall risk.
The hospital environment itself may either be a supportive environment (eg, the presence
of handrails and no-slip bathing surfaces) or may contribute to fall risk (eg, unfamiliar rooms,
improper bed height). This chapter reviews general evidence regarding multicomponent falls
prevention protocols, and 5 specific interventions: identification bracelets, physical restraints,
bed alarms, special flooring, and hip protectors.
Prevalence and Severity of the Target Safety Problem
Falls are among the most common incidents reported in institutions, 5 although incident
reports may underestimate their true occurrence. 6 The incidence of falls among hospitalized
patients varies according to the risk factors and case mix of the patient population as well as the
presence of falls prevention measures. Rubinstein has reported fall rates of 0.6 to 2.9 falls
annually per bed in hospitalized patients and 0.6 to 3.6 falls annually per bed in long-term care
institutions, based on published data. 7 About 50% of the 1.7 million nursing home residents in
the United States fall at least once each year, resulting in serious injury in about 10% of
residents. 7-9 The total cost of falls injuries in 1994 for adults aged 65 years and older was
estimated at $20.2 billion. 10
Hip fractures are the most feared complication of falls. Up to 20% of people sustaining a
hip fracture become nonambulatory, and only 14-21% recover their ability to carry out
281

instrumental activities of daily living. 11 The estimated total incremental costs (the difference
between costs before and after a hip fracture) of caring for an individual in the year after fracture
were estimated to be between $16,300 and $18,700. 12 Estimated Medicare expenditures for hip
fractures in 1991 were about $2.9 billion. 13
Practice Description
Based on the multifactorial etiology of falls, multicomponent interventions have been
developed to address patient risk factors and decrease fall rates. However, most studies have not
been designed in a way to determine which components of a multicomponent intervention are
most effective.
Risk Assessment
A variety of institution-based programs have been implemented to prevent falls. These
programs usually begin by identifying individuals at increased risk for falling. This is
accomplished by history-taking to elicit past falls history or by using more formal assessment
tools. 4,14-17 Protocols used to perform falls risk assessment in hospitals or nursing homes vary by
institution and often have not been validated. 18
Community-Dwelling Elders
An overwhelming majority of the large, prospective, controlled studies have been carried
out in the outpatient environment. They deserve mention because many of the interventions
could be modified for a hospital-based intervention. Tinetti and colleagues 19 showed that
interventions to reduce specific risk factors resulted in a 30% reduction in falls over one year in a
prospective community cohort. The targeted risk factors were postural hypotension, use of any
benzodiazepine or sedative-hypnotic drug, use of 4 or more prescription medications,
environmental hazards, and muscular strength or range of motion impairments. Specific
interventions that were part of the multicomponent program included exercise recommendations,
behavioral recommendations, medication review, and environmental modifications. A systematic
review of predominantly non-hospital based multi-risk factor intervention studies showed
significant protection against falling (Peto OR 0.77, 95% CI: 0.64-0.91). 20 There was, however,
significant heterogeneity across studies.
The large literature on community-based interventions has yielded other insights, some of
which may be applicable to the acute care setting. For example, exercise-based interventions 21-25
have been studied as a means to decrease falls in older persons. Results of these trials have not
been conclusive. A pre-planned meta-analysis of 7 randomized controlled trials (2 nursing homebased
and 5 community-based) that included an exercise component found a 10% decrease in
fall risk (adjusted incidence ratio 0.90, 95% CI: 0.81-0.99), 26 although a recent systematic review
examining the effect of 4 trials of exercise alone found no protection against falling. 20 Another
important insight from primarily non-hospital settings includes the association between specific
medications or classes of medications and falls. 27,28 Although several studies have used
pharmacist- or physician-based medication reviews as part of a multifaceted intervention, the
independent effect of medication review and adjustment on fall outcomes has not been reported.
Institutionalized Elders
In a nursing home setting, a promising randomized controlled trial incorporating
individualized assessment and targeting 4 falls-associated domains has been reported. 29
Intervention facilities had 19% fewer recurrent falls (95% CI: 2%-36%) compared with control
282

facilities and a 31% reduction in mean rate of injurious falls (13.7 vs. 19.9 falls per 100 personyears;
p=0.22). Interventions in this study were made in the areas of environmental and personal
safety (improvement in room lighting, flooring, footwear), wheelchair use and maintenance
(assessment by an occupational therapist), psychotropic drug prescription (assessment and
recommendations for change), transfer and ambulation (evaluation and recommendations for
change), and facility-wide interventions (eg, in-service educational programs). No analogous
study of a multi-intervention standardized protocol has been reported in hospitalized patients.
Hospitalized Elders
In the hospital, interventions have been employed as part of multiple risk factor
intervention studies, but many have been poorly described and standardized. In the studies set in
acute care environments, 30-45 practices include educational activities for nurse and support staff,
patient orientation activities, review of prior falls, and improvement of surrounding environment.
Specific environmental components included decreasing ward or room obstacles, adding
supplemental lighting and grab bars in bathrooms, and lowering bedrails and bed height. Other
studies have attempted to improve transfer and mobility by providing scheduled ambulatory and
physical therapy activities and better footwear (eg, non-skid socks). Additionally, studies have
incorporated strategies to assist cognitively impaired patients by educating family members to
deal with confused patients, minimizing sedating medications, and moving confused patients
closer to nursing staff. Because many of these hospital studies use small sample sizes and
inadequately describe the precise number and standardization of interventions, their
generalizability and reproducibility is limited. However, a recent systematic review of many of
these programs concluded that a pooled effect of 25% reduction in the fall rate occurred in the
studies that examined prospective interventions compared to fall risk in historical controls. 18
Some interventions with the potential for effectiveness in isolation have been studied.
Each of the following hospital- or institution-based individual interventions has been analyzed
independently of a multi-component falls prevention program:
• Identification bracelets
• Physical restraints
• Bed alarms
• Special flooring
• Hip protectors
Several generally accepted interventions with high face-validity have not been
independently studied, yet are commonly accepted practices. Immobility 46 is a significant risk
factor for several geriatric complications, including falls, pressure ulcers, and functional decline.
Minimization of bedrest is a practical, real-world intervention that has implications for
prevention of a number of serious hospital-acquired complications. 47
Comment
There are few hospital or other institution-based randomized controlled trials of
standardized falls interventions, although the necessity for well-designed studies is clear. The
nursing home-based intervention reported by Ray and colleagues 29 provides good evidence that a
well-documented intervention can improve falls outcomes in institutionalized patients. No
similarly designed trial of a multicomponent intervention in hospitalized patients was identified,
although many falls prevention programs incorporate multifactorial interventions. The questions
raised by multicomponent falls prevention studies include the generalizability of interventions to
283

diverse inpatient settings, appropriate targeting of at-risk individuals, analysis of the individual
components that provide the best improvement in falls outcomes, and the transportability of
interventions between institutions with variable resources for implementation. Evidence for the
effectiveness of individual interventions is important, but effectiveness may change (for better or
worse) when such interventions are incorporated with others as part of a falls prevention
program.
Subchapter 26.1. Identification Bracelets for High-Risk Patients
Background
Some hospitals use colored bracelets to identify patients at high risk for falls. Other
identification methods include signs, stickers, or tags placed above the patient’s bed, at the
nursing station, or on the patient’s chart. In theory, these remind staff that the patient is at high
risk for falls and trigger interventions that reduce the risk of falls (eg, supervision or assistance
with ambulation, minimization of sedative-hypnotic medications, lowering of bed height).
Identification bracelets might also impact patients’ falls awareness (eg, reminding patients to call
for assistance before getting out of bed).
Prevalence and Severity of the Target Safety Problem
See Introduction to Chapter 26.
Opportunities for Impact
We found no published data on the number of hospitals currently using such strategies.
Practice Description and Evidence for Effectiveness
A search of the literature identified many studies that have used identification bracelets,
signs, or tags for high-risk patients. 31-33,35,40-42,44,45,48,49 Most of these involved multiple,
simultaneous interventions and were designed such that estimation of the treatment effect due to
the identification bracelet, signs or tags component cannot be calculated. The remaining study
was a randomized, controlled trial of colored identification bracelets worn by inpatients at high
risk for falls (Table 26.1.1). 50 “High-risk” was defined as history of multiple falls, an episode of
incontinence, or an admitting diagnosis of stroke or ataxia. Cox proportional hazards model was
used to assess the effect of identification bracelets on time-to-first-fall. The fall rate was 42%
(27/65) in the intervention group and 30% (21/69) in the control group, which did not represent a
statistically significant difference. After preliminary analysis of the data, the investigators and
ethics committee agreed that it was not appropriate to continue for the sole purpose of obtaining
statistical power, and the study was terminated.
Potential for Harm
None identified.
Costs and Implementation
Identification tags and similar interventions are associated with minimal costs.
Comment
Use of special bracelets, signs, and stickers to identify patients at high risk for falls is a
relatively inexpensive and easy to implement practice. There is currently insufficient information
284

as to whether identification bracelets, as a isolated intervention, decrease falls. Future studies
should assess the effectiveness of similar identification strategies in the context of
multicomponent fall prevention programs and, if they are effective, which methods work best.
Table 26.1.1. Study of identification bracelets*
Study Participants and Setting Study Design,
Outcomes
Mayo,
1994 50
134 high-risk patients in a
rehabilitation hospital, 1990-91
* CI indicates confidence interval.
References
Level 1,
Level 1
285
Results
Hazard ratio for fall with
intervention: 1.3 (95% CI: 0.8-2.4)
1. Baker SP, O'Neill B, Ginsburg MJ, Guohua L. The injury fact book. 2nd edition. New
York: Oxford University Press, 1992.
2. Tinetti ME, Speechley M. Prevention of falls among the elderly. N Engl J Med.
1989;320:1055-9.
3. Myers AH, Young Y, Langlois JA. Prevention of falls in the elderly. Bone. 1996;18:87S-
101S.
4. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living
in the community. N Engl J Med. 1988;319:1701-1707.
5. Sutton J, Standan P, Wallace A. Incidence and documentation of patient accidents in
hospital. Nurs Times. 1994;90:29-35.
6. Kanten DN, Mulrow CD, Gerety MB, Lichtenstein MJ, Aguilar C, Cornell JE. Falls: an
examination of three reporting methods in nursing homes. J Am Geriatr Soc.
1993;41:662-666.
7. Rubenstein LZ, Robbins AS, Schulman BL, Rosado J, Osterweil D, Josephson KR. Falls
and instability in the elderly. J Am Geriatr Soc.1988;36:266-278.
8. Rubenstein LZ, Josephson KR, Robbins AS. Falls in the nursing home. Ann Intern Med.
1994;121:442-451.
9. Tinetti ME. Factors associated with serious injury during falls by ambulatory nursing
home residents. J Am Geriatr Soc. 1987;35:644-648.
10. Englander F, Hodson TJ, Terregrossa RA. Economic dimensions of slip and fall injuries.
J Forensic Sci. 1996; 41:733-746.
11. Zuckerman JD. Hip fracture. N Engl J Med. 1996;334:1519-1525.
12. Brainsky GA, Lydick E, Epstein R, et al. The economic cost of hip fractures in
community-dwelling older adults: a prospective study. J Am Geriatr Soc.1997;45:281-
287.
13. CDC. Incidence and costs to Medicare of fractures among Medicare beneficiaries aged >
65 years: United States, July 1991-June 1992. MMWR - Morbidity & Mortality Weekly
Report 1996;45:877-883.
14. Oliver D, Britton M, Seed P, Martin FC, Hopper AH. Development and evaluation of
evidence based risk assessment tool (STRATIFY) to predict which elderly inpatients will
fall: case-control and cohort studies. BMJ. 1997;315:1049-1053.
15. Whitney SL, Poole JL, Cass SP. A review of balance instruments for older adults. Am J
Occup Ther. 1998;52: 666-671.

16. Shumway-Cook A, Baldwin M, Polissar NL, Gruber W. Predicting the probability for
falls in community-dwelling older adults. Phys Ther. 1997;77:812-819.
17. Tinetti ME. Performance-oriented assessment of mobility problems in elderly patients. J
Am Geriatr Soc. 1986;34:119-126.
18. Oliver D, Hopper A, Seed P. Do hospital fall prevention programs work? A systematic
review. J Am Geriatr Soc. 2000;48:1679-1689.
19. Tinetti ME, Baker DI, McAvay G, et al. A multifactorial intervention to reduce the risk
of falling among elderly people living in the community. N Engl J Med. 1994;331:821-
827.
20. Gillespie LD, Gillespie WJ, Cumming R, Lamb SE, Rowe BH. Interventions for
preventing falls in the elderly. In: The Cochrane Library, Issue 2, 2000. Oxford: Update
Software.
21. McMurdo ME, Mole P, Paterson C. Controlled trial of weight-bearing exercise in older
women in relation to bone density and falls. BMJ. 1997;314:569.
22. Wolf SL, Barnhart HX, Kutner NG, McNeely E, Coogler C, Xu T. Reducing frailty and
falls in older persons: an investigation of Tai Chi and computerized balance training.
Atlanta FICSIT Group. Frailty and Injuries: Cooperative Studies of Intervention
Techniques. J Am Geriatr Soc.1996;44:489-497.
23. Lord SR, Ward JA, Williams P, Strudwick M. The effect of a 12-month exercise trial on
balance, strength, and falls in older women: a randomized controlled trial. J Am Geriatr
Soc. 1995;43:1198-1206.
24. Reinsch S, MacRae P, Lachenbruch PA, Tobis JS. Attempts to prevent falls and injury: a
prospective community study. Gerontologist. 1992;32:450-456.
25. Mulrow CD, Gerety MB, Kanten D, et al. A randomized trial of physical rehabilitation
for very frail nursing home residents. JAMA. 1994;271:519-524.
26. Province MA, Hadley EC, Hornbrook MC, et al. The effects of exercise on falls in
elderly patients. A preplanned meta-analysis of the FICSIT Trials. Frailty and Injuries:
Cooperative Studies of Intervention Techniques. JAMA. 1995;273:1341-1347.
27. Leipzig RM, Cumming RG, Tinetti ME. Drugs and falls in older people: a systematic
review and meta-analysis: I. Psychotropic drugs. J Am Geriatr Soc.1999;47:30-39.
28. Leipzig RM, Cumming RG, Tinetti ME. Drugs and falls in older people: a systematic
review and meta-analysis: II. Cardiac and analgesic drugs. J Am Geriatr Soc. 1999;47:40-
50.
29. Ray WA, Taylor JA, Meador KG, et al. A randomized trial of a consultation service to
reduce falls in nursing homes. JAMA. 1997;278:557-562.
30. Rainville NG. Effect of an implemented fall prevention program on the frequency of
patient falls. Qual Rev Bull. 1984;10:287-291.
31. Fife DD, Solomon P, Stanton M. A risk/falls program: code orange for success...geriatric
patients. Nurs Manage. 1984;15:50-53.
32. Barker SM, O'Brien CN, Carey D, Weisman GK. Quality improvement in action: a falls
prevention and management program. Mt Sinai J Med. 1993;60:387-90.
33. Hendrich AL. An effective unit-based fall prevention plan. J Nurs Qual Assur.
1988;3:28-36.
34. Craighead J, Fletcher P, Maxwell J. Seven steps for fall prevention. Dimens Health Serv.
1991;68:25-26.
35. Kilpack V, Boehm J, Smith N, Mudge B. Using research-based nursing interventions to
decrease patient falls. Appl Nurs Res. 1991;4:50-56.
286

36. Mitchell A, Jones N. Striving to prevent falls in an acute care setting–action to enhance
quality. J Clin Nurs. 1996;5:213-220.
37. Hill BA, Johnson R, Garrett BJ. Reducing the incidence of falls in high risk patients. J
Nurs Adm. 1988;18:24-28.
38. Hernandez M, Miller J. How to reduce falls. Geriatr Nurs. 1986;7:97-102.
39. Krishna KM, Van Cleave RJ. Decrease in the incidence of patient falls in a geriatric
hospital after educational programs [letter]. J Am Geriatr Soc.1983;31:187.
40. Morton D. Five years of fewer falls. Am J Nurs. 1989;89:204-205.
41. Sweeting HL. Patient fall prevention–a structured approach. J Nurs Manag. 1994;2:187-
92.
42. Schmid NA. Reducing patient falls: a research-based comprehensive fall prevention
program. Mil Med. 1990;155:202-207.
43. Brady R, Chester FR, Pierce LL, Salter JP, Schreck S, Radziewicz R. Geriatric falls:
prevention strategies for the staff. J Gerontol Nurs. 1993;19:26-32.
44. Cannard G. Falling trend. Nurs Times. 1996;92:36-37.
45. Zepp S. "Ban a fall." A nursing innovation to reducing patient falls. Kans Nurse.
1991;66:13.
46. Mahoney JE. Immobility and falls. Clin Geriatr Med.1998;14:699-726.
47. Harper CM, Lyles YM. Physiology and complications of bed rest. J Am Geriatr Soc.
1988;36:1047-1054.
48. Rogers S. Reducing falls in a rehabilitation setting: a safer environment through team
effort. Rehabil Nurs. 1994;19:274-276,322.
49. Innes E. Maintaining fall prevention. QRB. 1985;11:217-221.
50. Mayo NE, Gloutney L, Levy AR. A randomized trial of identification bracelets to
prevent falls among patients in a rehabilitation hospital. Arch Phys Med Rehabil.
1994;75:1302-1308.
Subchapter 26.2. Interventions that Decrease the Use of Physical Restraints
Background
The Health Care Financing Administration (HCFA) defines physical restraints as “any
manual method or physical or mechanical device, material, or equipment attached or adjacent to
the patient that the individual cannot remove easily which restricts freedom of movement or
normal access to one’s body.” 1 Physical restraints have been used in nursing homes and hospitals
both as a safety device and as a falls prevention tool. Because restrained patients cannot arise
from a chair or transfer out of bed, they theoretically will not fall or, in the case of bedrails, will
not roll out of bed. However, the use of physical restraints may lead to substantial adverse
outcomes. In fact, serious injuries and even death have been reported with use of these devices. 2,3
This chapter examines interventions to reduce use of physical restraints and the concomitant
effect on fall rates.
Practice Description
Studies examining the use physical restraints have considered 2 types of interventions in
hospital or nursing home settings: bedrails and other mechanical restraints designed to restrict
mobility. These interventions usually begin with either a physician or nurse making an
assessment that a patient is at risk for falls, elopement, or other adverse outcomes. Thereafter,
287

use of a restraint is initiated, with periodic reassessment of the ongoing need for the device.
Safety practices to reduce restraint use in nursing home patients have included nursing education
strategies focusing on assessment/reassessment of the need for restraints and the use of
alternatives to restraints.
Prevalence and Severity of the Target Safety Problem
See Introduction to Chapter 26.
Opportunities for Impact
Federal guidelines now discourage all but the limited, appropriate use of physical
restraints and bedrails. Legislation adopted as part of the Omnibus Budget Reconciliation Act of
1987 directed nursing homes to limit physical restraints, and the Joint Commission on the
Accreditation of Healthcare Organizations (JCAHO) has adopted similar guidelines. Several
statewide initiatives (eg, the Pennsylvania Restraint Reduction Initiative, begun in 1996) have
been implemented under HCFA’s National Restraint Reduction Initiative, resulting in notable
reductions in restraint usage. 4 The Food and Drug Administration’s Hospital Bed Safety Work
Group has likewise actively raised awareness of the risks and benefits of bedrail use. 5 Based on
an annual HCFA survey, the national restraint rate was approximately 13.5% in 1999, down
from approximately 20% in 1996 when HCFA’s Restraint Reduction Initiative began. 6
Nonetheless, data from selected states reveals that the rate was still as high as 26% as of 1998. 7
Study Designs and Outcomes
Six studies were identified: 2 concerning bedrail interventions 8,9 and 4 describing
mechanical restraints interventions (Table 26.2.1). 7,10-12 Most studies compare interventions with
historical control or baseline rates using a before-after study design. Morbidity data on falls are
reported in all studies.
Evidence for Effectiveness of the Practice
The studies reveal no statistically significant difference in falls compared with historical
controls when bedrails are removed. In fact, restrained patients appear to have a modest increase
in fall risk or fall injuries based on several studies. Weaknesses in study design for some of these
studies preclude a final conclusion.
Potential for Harm
The potential for harm with use of bedrails is well-documented, including death from a
variety of mechanisms, including death and strangulation. 13 Mechanical restraints likewise carry
a risk of severe injury, strangulation, and mobility limitations that may predispose patients to
other adverse outcomes (pressure ulcers, incontinence, acute confusion). Limits to patient
freedom, dignity, and quality of life also contribute to the potential for harm. A potential harm of
interventions to decrease restraint use is that there may be an increase in other adverse events
(eg, elopement) if appropriate alternative preventive measures are not in place.
Costs and Implementation
The costs associated with interventions to reduce the use of restraints have not been
described. Nonetheless, reduction in the use of physical restraints will require resources to pay
for alternative interventions and rehabilitative measures and will increase labor costs. 14
Compliance with interventions to reduce bedrail rates and to decrease mechanical restraint use
288

has been good. In fact, given adequate alternatives to the use of these devices, hospital and
nursing staffs have decreased their usage significantly. In the Neufeld study, 7 for example,
restraint use fell from 41% to 4%.
Comment
There is growing evidence that physical restraints have a limited role in medical care.
Restraints limit mobility, a shared risk factor for a number of adverse geriatric outcomes, and
increase the risk of iatrogenic events. They certainly do not eliminate falls, and decreasing their
use can be accomplished without increasing fall rates. In some instances reducing the use of
restraints may actually decrease the risk of falling. Incorporating changes into physician and staff
behavior may be easier if large, multicenter trials are successful in identifying safe alternatives to
restraints that effectively limit falls risks for patients.
289

Table 26.2.1. Studies of physical restraints and fall risk*
Study Participants and Setting Design,
Outcomes
Hanger,
1999 9
1968 hospital patients in New
Zealand, 1994; formal bedrail
policy and educational
program to reduce bedrail use,
historical controls
Si, 1999 8 246 patients in a teaching
nursing home, 1993-94;
interdisciplinary team
assessment and removal of
bedrails with provision of
bedrail alternatives, historical
controls
Capezuti,
1996 11
Capezuti,
1998 12
Neufeld,
1999 7
Tinetti,
1991 10
322 nursing home residents;
subgroup of confused patients
examined for mechanical
restraint use
633 nursing home residents in
3 nursing homes, 1990-1991;
restraint education and
consultation interventions
compared with baseline rates
2075 nursing home beds in 16
nursing homes, 1991-1993;
educational intervention to
decrease mechanical restraints
compared with baseline rates
397 elderly patients at 12
skilled nursing facilities;
observational cohort study of
mechanical restraint use
Level 3,
Level 1
Level 3,
Level 1
Level 3,
Level 1
Level 3,
Level 1
Level 3,
Level 1
Level 3,
Level 1
* CI indicates confidence interval; OR, odds ratio.
290
Results
No significant difference in overall fall
rate: 164.8 falls/10,000 bed days
before and 191.7 falls/10,000 bed
days after the intervention (p=0.18)
Fewer serious falls occurred after the
intervention (p=0.008)
No significant difference in fall rates:
2/116 (1.7%) patients before and
2/130 (1.5%) patients after the
intervention
Confused patients who were restrained
had increased odds of falling (OR
1.65, 95% CI: 0.69-3.98) and
recurrent falls (OR 2.46, 95% CI:
1.03-5.88)
No significant increase in fall rates in
the restraint-free group
Decreased odds of minor injury after
restraint removal, adjusted OR 0.3
(95% CI: 0.1-0.9)
Moderate/severe injuries decreased
from 7.4% to 4.4% (p=0.0004) after
educational intervention
15/275 (5%) of unrestrained patients
compared to 21/122 (17%)
experienced a serious fall-related
injury (p

References
1. Health Care Financing Administration. FY 2001 Annual Performance Plan. Available at:
http://www.hcfa.gov/stats/2001.htm. Accessed June 13, 2001.
2. FDA Center for Devices and Radiological Health. FDA safety alert: entrapment hazards
with hospital bed side rails. August 1995. Available at: http://www.fda.gov/cdrh/
bedrails.html. Accessed June 12, 2001.
3. Miles SH, Irvine P. Deaths caused by physical restraints. Gerontologist. 1992;32:762-766.
4. Health Care Financing Administration. National restraint reduction newsletter. Summer
2000. Available at: http://www.hcfa.gov/publications/newsletters/restraint/2000/
rr0600.asp. Accessed June 12, 2001.
5. FDA Center for Devices and Radiological Health. A guide to bed safety. Available at:
http://www.fda.gov/cdrh/beds/bedrail.pdf. Accessed June 12, 2001.
6. Health Care Financing Administration. National restraint reduction newsletter. Winter
1999. Available at: http://www.hcfa.gov/publications/newsletters/restraint/1999/
rrwin99.htm. Accessed June 12, 2001.
7. Neufeld RR, Libow LS, Foley WJ, Dunbar JM, Cohen C, Breuer B. Restraint reduction
reduces serious injuries among nursing home residents. J Am Geriatr Soc. 1999;47:1202-
1207.
8. Si M, Neufeld RR, Dunbar J. Removal of bedrails on a short-term nursing home
rehabilitation unit. Gerontologist. 1999;39:611-614.
9. Hanger HC, Ball MC, Wood LA. An analysis of falls in the hospital: can we do without
bedrails? J Am Geriatr Soc. 1999;47:529-531.
10. Tinetti ME, Liu YB, Ginter S. Mechanical restraint use and fall related injuries among
residents of skilled nursing facilities. Ann Intern Med. 1992;116:369-374.
11. Capezuti E, Evans L, Strumpf N, Maislin G. Physical restraint use and falls in nursing
home residents. J Am Geriatr Soc. 1996;44:627-633.
12. Capezuti E, Strumpf NE, Evans LK, Grisso JA, Maislin G. The relationship between
physical restraint removal and falls and injuries among nursing home residents. J Gerontol
A Biol Sci Med Sci. 1998;53:M47-M52.
13. Parker K, Miles SH. Deaths caused by bedrails. J Am Geriatr Soc. 1997;45:797-802.
14. Schnelle JF, Smith RL. To use physical restraints or not? [editorial]. J Am Geriatr Soc.
1996;44:727-728.
Subchapter 26.3. Bed Alarms
Background
Epidemiologic studies reveal that falls occur commonly in and around bed areas. 1,2
Decreasing the risk of falls when patients attempt to transfer into and out of bed without
assistance is a potentially important target safety goal. This chapter examines the use of a bed
alarm system that alerts hospital staff to patient movement out of bed as a strategy to reduce
falls. General principles of alarm use in health care settings can be found in Chapter 8.
291

Practice Description
A sensor device is placed on the bed, under a sitting or reclining patient. When the
patient changes position, it detects movement and/or absence of weight. An audible alarm is
triggered at the nurses’ station and, with some devices, in the patient’s room. The alarm alerts
nurses when patients attempt to leave the bed without assistance and may alert a patient to
remain in bed if the alarm is audible in the patient’s room.
Evidence for Effectiveness of the Practice
Several studies have included bed alarms as part of a multifaceted intervention. 3-6
However, the study designs do not allow calculation of the effect attributable to the bed alarm
component or were not controlled. A recent, unpublished before-after study was identified in a
Web search but the full report could not be obtained before completion of this chapter. 7
Tideiksaar et al randomized elderly patients at “high risk” for falls to either a group that received
an alarm system (the RN+ OnCall bed monitoring system) or to a control group that did not
(Table 26.3.1) 8 . The groups were similar in age and gender. No other baseline comparisons were
reported. There were fewer falls in the study group but the difference failed to reach statistical
significance. However, the total number of falls was low (n=17) and had there been one less fall
in the alarm group or one more fall in the control group, the difference would have been
statistically significant.
Potential for Harm
No harm was identified. There are theoretical electrical risks if the sensor devices are
internally compromised due to bending of the sensor mats and exposure to fluids, but such
events have not been reported in the literature.
Costs and Implementation
Costs of the devices vary by manufacturer, the type of bed monitoring system used, and
the number of beds to be monitored. Manufacturers’ charges range from several hundred to
several thousand dollars for the receiving equipment. Individual sensors require replacement
after pre-specified periods of use or, in some cases, can be cleaned between patients, which
incurs additional hospital costs. Implementation requires adequate staffing to respond in a timely
manner to the audible alarms.
Comment
At this time, there is insufficient evidence regarding the effectiveness of bed alarms in
preventing falls in elderly patients to recommend the practice. Additional research sufficiently
powered to identify meaningful differences, coupled with a formal economic analysis, would be
useful.
292

Table 26.3.1. Study of bed alarms for fall prevention
Study Participants and Setting Study Design,
Outcomes
Tideiksaar,
1993 8
70 patients on a geriatric unit in
a university hospital, 1992
Level 1,
Level 1
293
Results
(95% Confidence Interval)
Odds ratio for prevention of
falls: 0.32 (0.10-1.03)
References
1. Vassallo M, Amersey RA, Sharma JC, Allen SC. Falls on integrated medical wards.
Gerontology. 2000;46:158-162.
2. Sutton J, Standan P, Wallace A. Incidence and documentation of patient accidents in
hospital. Nurs Times. 1994;90:29-35.
3. Schmid NA. Reducing patient falls: a research-based comprehensive fall prevention
program. Mil Med. 1990;155:202-207.
4. Widder B. A new device to decrease falls. Geriatr Nurs. 1985;6:287-288.
5. Morton D. Five years of fewer falls. Am J Nurs. 1989;89:204-205.
6. Innes E. Maintaining fall prevention. QRB. 1985;11:217-221.
7. Bed-Check Corporation. Fall prevention study: bed alarms, investigating their impact on
fall prevention and restraint use. Available at: http://www.bedcheck.com/fall-preventionstudy.html.
Accessed June 6, 2001.
8. Tideiksaar R, Feiner CF, Maby J. Falls prevention: the efficacy of a bed alarm system in an
acute-care setting. Mt Sinai J Med. 1993;60:522-527.
Subchapter 26.4. Special Hospital Flooring Materials to Reduce Injuries from Patient
Falls
Background
One proposed practice to prevent injury due to falls is to alter flooring material on
hospital wards or in nursing homes. Carpeting, vinyl, or other biomedically-engineered materials
could potentially improve falls outcomes. The use of special flooring materials has been shown
to influence specific gait characteristics in hospitalized elders. 1 A recent study described the
Penn State Safety Floor, which is designed to remain relatively rigid under normal walking
conditions but to deform elastically to absorb impact forces during a fall. 2 The efficacy of this
floor is still being tested outside the laboratory environment among nursing home residents. 3
Practice Description
As data on the efficiency of the Penn State Safety Floor 2 are not yet available, we restrict
our review to the use of hospital-duty carpeting compared with “usual” vinyl flooring.

Study Designs and Outcomes
We identified 2 studies of the effect of flooring type (carpet vs. “usual” vinyl flooring) on
falls: a randomized controlled trial in an inpatient rehabilitation unit 4 and a retrospective study of
accidents reported in a care of the elderly unit (Table 26.4.1). 5 Both studies reported Level 1
outcomes. The randomized trial measured the rate of falls. The retrospective analysis studied
fall-related injury, defined as any graze, bruise, laceration, fracture or pain.
Evidence for Effectiveness of the Practice
The randomized trial by Donald et al found more falls in the group housed in rooms with
carpeted flooring, although the difference barely failed to achieve statistical significance. The
earlier retrospective analysis by Healey found that the rate of injury was significantly lower for
patients who fell on carpet rather than vinyl flooring. 5 The severity of injuries was not reported
and it was not possible to determine whether the rate of falls differed according to flooring
material.
Potential for Harm
No harm was identified, although it is possible that asthmatic patients might react to
increased levels of dust-mite allergens in carpeted wards. 6
Costs and Implementation
No cost estimates for changes in flooring were reported in the literature. Implementation
of this practice would require a large expenditure for facilities upgrades nationwide. Likewise,
the costs associated with keeping various floor surfaces clean in the hospital or nursing home
environment would also be high.
Comment
Advances in biomedical engineering could result in potentially significant redesign of the
physical environment in hospitals and nursing facilities. The primary aim of specialized flooring
could be either to reduce the risk of falling or to reduce the risk of an injury once a fall has
occurred, or both. The two studies analyzed seem to indicate that carpeted floors may increase
fall rates but decrease fall injuries; it is possible that other surfaces would yield better results.
Further study of this area is warranted.
294

Table 26.4. Study of special flooring for falls prevention
Study Participants and Setting Study Design,
Outcomes
Donald,
2000 4
Healey,
1994 5
32 patients in an elderly care
rehabilitation ward in the
United Kingdom in 1996
Random sample of accident
forms (n=213) from care of
elderly unit over 4 years
Level 2,
Level 1
Level 3,
Level 1
295
Results
Rate of falls:
Carpet, 10/16 (63%); vinyl, 1/16 (6%)
RR 8.3 (95% CI: 0.95-73; p=0.05)
Falls resulting in injury:
Carpet, 15%; vinyl, 91% (p

Practice Description
External hip protectors are usually made with plastic pads or shields that are padded or
constructed with foam-type materials. They fit into specially-designed pockets in undergarments
or pants. They are designed to be worn during the day for people who are out of bed, walking or
engaged in activities that place them at higher risk for falls. Ideally, they would be worn all the
time to protect individuals from nighttime falls.
Prevalence and Severity of the Target Safety Problem
See Introduction to Chapter 26.
Opportunities for Impact
No data on the nationwide use of hip protectors in the hospital or nursing home are
available. A small minority of institutions are in the process of evaluating them, and a few may
have begun to use them.
Study Designs
Five relevant randomized controlled trials 4-8 were identified from a literature search and
from a Cochrane systematic review. 9 The Cochrane review cites 2 additional abstracts 10,11 not
included here. Four of the trials evaluate effectiveness of the devices and one study 8 examines
compliance rates of wearing hip protectors as part of a pilot study. Two studies were clusterrandomized
and 2 were randomized by individual patient.
Study Outcomes
Studies reported hip fractures as an outcome, although compliance with the intervention
was the primary outcome in one study. Additional outcomes reported were mortality, falls, and
non-hip fractures.
Evidence for Effectiveness of the Practice
External hip protectors appear to be an effective means to reduce the risk of a hip fracture
in older persons aged 65 and over who fall. Table 26.5.1 lists the abstracted studies and outlines
their pertinent features. The generalizability of these results to wider audiences and to lower risk
populations has not been demonstrated, nor has the potential benefit for hospitalized patients
been reported. Concerns with compliance could hinder their effectiveness on a population-wide
level.
Potential for Harm
Discomfort from wearing the device, difficulty managing the garment while dealing with
continence, and the potential for skin irritation and breakdown are causes for concern if fragile
older people were to wear hip protectors. Because long-term compliance is low, it is unclear how
many people would experience such problems if the devices were worn for longer periods during
the day or for long-term use.
296

Costs and Implementation
An Australian study published in 2000 quoted a cost of A$10 per pair (approximately
$5.25US). 4 The retail price quoted by one US manufacturer of a different hip protector is
approximately $90 per pair. The lycra-containing undergarment used by some manufacturers to
keep the hip pads in place requires special laundering and would require a tracking system
similar to that used for other specialized garments or medical devices assigned to patients within
a facility. Once provided, if devices can be put on and taken off by individual users,
implementation is straightforward. The cost-effectiveness of the devices has not been formally
reported.
Comment
One of the main philosophical concerns raised by these studies is the change in emphasis
from primary prevention of the underlying cause of hip fractures (ie, falls) to an emphasis on
methods of protecting patients from the deleterious consequences of falls. However, a strategy
for addressing the multiple risk factor model for falls is still warranted for primary falls
prevention. With this caveat in mind, there is strong evidence to support the ability of hip
protectors to prevent hip fractures. This evidence, in addition to their high face validity, may
encourage their rapid adoption. Further evaluation of their costs, acceptability to patients, and
effectiveness in hospitalized patients (versus nursing home residents) is needed.
297

Table 26.5.1. Hip protectors to prevent hip fracture*
Study Participants and
Setting
Parker,
2000 9
Chan,
2000 4
Ekman,
1992 5
Kannus,
2000 6
Lauritzen,
1993 7
Villar,
1998 8
1752 nursing home or
rest home residents in
5 countries
71 nursing home
residents in Australia,
year not stated
746 nursing home
residents in Sweden,
year not stated
1801 community
based elderly in
Finland, 1996-1997
665 nursing home
residents in Denmark,
1991-1992
141 rest home
residents in the UK,
year not stated
Design,
Outcomes
Level 1A,
Level 1
Level 1,
Level 1
Level 1,
Level 1
Level 1,
Level 1
Level 1,
Level 1
Level 1,
Level 3
298
Results
* CI indicates confidence interval; OR, odds ratio; and RR, relative risk.
References
Peto OR 0.44 (95% CI: 0.26-0.75) of hip
fracture in the intervention group in clusterrandomized
studies;
Peto OR 0.22 (95% CI: 0.09-0.57) in patientrandomized
studies
RR of hip fracture in the intervention group
0.264 (95% CI: 0.073-0.959)
RR of hip fracture in the intervention group
0.33 (95% CI: 0.11-1.00)
RR of hip fracture in the intervention group
0.4 (95% CI: 0.2-0.8; p=0.008)
RR of hip fracture in the intervention group
0.44 (95% CI: 0.21-0.94)
30% compliance over 3 months (hip fracture
outcomes not assessed)
1. Stevens JA, Olson S. Reducing falls and resulting hip fractures among older women.
MMWR Morb Mortal Wkly Rep. 2000;49:1-12.
2. Hannan EL, Magaziner J, Wang JJ, et al. Mortality and locomotion 6 months after
hospitalization for hip fracture. JAMA. 2001;285:2736-2742.
3. Magaziner J, Hawkes W, Hebel JR, et al. Recovery from hip fracture in eight areas of
function. J Gerontol. Medical Sciences 2000;55A:M498-M507.
4. Chan DK, Hillier G, Coore M, et al. Effectiveness and acceptability of a newly designed
hip protector: a pilot study. Arch Gerontol Geriatr. 2000;30:25-34.
5. Ekman A, Mallmin H, Michaelsson K, Ljunghall S. External hip protectors to prevent
osteoporotic hip fractures. Lancet. 1997;350:563-564.
6. Kannus P, Parkkari J, Niemi S, et al. Prevention of hip fracture in elderly people with use
of a hip protector. N Engl J Med. 2000;343:1506-1513.

7. Lauritzen JB, Petersen MM, Lund B. Effect of external hip protectors on hip fractures.
Lancet 1993;341:11-13.
8. Villar MTA, Hill P, Inskip H, Thompson P, Cooper C. Will elderly rest home residents
wear hip protectors? Age Ageing. 1998;27:195-198.
9. Parker MJ, Gillespie LD, Gillespie WJ. Hip protectors for preventing hip fractures in the
elderly. In: The Cochrane Library, Issue 2, 2001. Oxford: Update Software.
10. Harada A, Okulzumi H. Hip fracture prevention trial using hip protector in Japanese
elderly. Osteoporos Int. 1998;8(Suppl 3):121.
11. Heikinheimo R, Jantti PL, Aho HJ, Maki-Jokela PL. To fall but not to break – safety pants.
Proceedings of the Third International Conference on Injury Prevention and Control,
Melborne 1996;1996:576-578.
299

300

Chapter 27. Prevention of Pressure Ulcers in Older Patients
Joseph V. Agostini, MD
Dorothy I. Baker, PhD, RNCS
Sidney T. Bogardus, Jr., MD
Yale University Schools of Medicine and Public Health
Background
Pressure ulcers, localized areas of tissue damage or necrosis that develop due to pressure
over a bony prominence, are common causes of morbidity in older hospitalized and
institutionalized persons. Other terms referring to the same phenomena are pressure sores,
bedsores, and decubitus ulcers. Risk factors include immobility, friction, shear, incontinence,
cognitive impairment and poor nutritional status. 1-3 Pressure ulcers are one indicator of quality of
care measured by nursing homes as part of the mandatory Minimum Data Set (MDS), which is
required for participation in Medicare and Medicaid. Part of the MDS evaluation includes the
Resident Assessment Instrument, which serves as a guide to assess pressure ulcers and many
other pertinent clinical problems. 4
Risk assessment is an integral part of prevention efforts. The Norton scale 5 and the
Braden scale 6 are widely used tools to identify at-risk patients. The Norton scale assesses five
domains: activity, incontinence, mental status, mobility, and physical condition. The Braden
scale assesses six domains: activity, dietary intake, friction, mobility, sensory perception, and
skin moisture. Agreement between the scales is 0.73 using the kappa statistic. 7
Different strategies have been used for primary prevention. Major clinical guidelines, 8 for
pressure ulcer prevention are based primarily on published evidence, and in some areas, on
professional judgment and face validity of practices. Turning and/or repositioning patients is a
practice with high face validity, but there are no well-designed controlled trials that examine its
effect in the absence of other interventions. Other practices include regular skin inspection and
assessment, use of appropriate pressure-relief surfaces, improved mobility, adequate nutritional
intake, and documentation of the skin examination. Additionally, the use of general educational
interventions for hospital staff is supported by before-after study designs. 9, 10 Several reports
suggest the value of using topical applications applied to intact skin in an attempt to prevent
ulcers. 5, 11, 12 This chapter focuses on the use of pressure-relieving strategies that can be
incorporated into hospital or nursing home practice and are based on evidence from controlled
clinical trials.
Practice Description
The preventive practices that have received the most research attention are the use of
specific beds or mattresses. Many studies have compared a specific mattress with either another
high-technology mattress or with a “standard” hospital mattress. A standard mattress, which is
not uniformly defined in the literature, may be described as a hospital-issue, usually foam-based
mattress found in a typical hospital room. The lack of consensus as to what constitutes a standard
hospital mattress presents an interpretative challenge to investigators and administrators hoping
to extrapolate results of a published trial to another locale.
In 1991 Krasner reported that there were over 115 different pressure-relieving support
surfaces on the market. 13 Sheepskin and other inexpensive pads (eg, “egg-crate” mattresses) are
common pressure-reducing devices. Other static devices include pressure-relieving pads (eg,
301

fabricated from elastic polymers) such as those used to cover operating room tables. Constant,
low-pressure supports maintain uniform pressure throughout. Examples include higher-grade
foam, and gel-, air-, or water-filled supports. In contrast, dynamic or alternating air devices have
a built-in pump that continually redistributes air pressure. Low air-loss beds, as their name
implies, permit small amounts of air to escape through a network of pores, whereas high air-loss
(or air-fluidized) beds purposefully pump air through ceramic-type beads. Finally, kinetic
turning beds, which allow continual rotation, are used more commonly in critically ill patients.
Prevalence and Severity of the Target Safety Problem
In 1990 a large, prospective epidemiologic study reported the one-year incidence for
pressure ulcer development in nursing homes to be 13.2%, 14 with prevalence reports ranging
from 7% to 23% in a systematic review. 15 Risk-adjusted rates of new pressure ulcers have been
reported to decrease by 25% from 1991 to 1996, based on a recent study using data from the
Minimum Data Set. 16 In hospitalized patients, prevalence ranges from about 3% to 11%. 17
Meehan reported that the prevalence of pressure ulcers was 11.1% in 177 hospitals surveyed in
1993 and that 54% of the pressure ulcers occurred in patients 70 to 89 years old. 18 Eckman
estimated that almost 1.7 million hospitalized patients had pressure ulcers. 19 Approimately 60%
of pressure ulcers develop in these acute care settings. The National Health and Nutrition
Examination Survey found that less than 20% of pressure ulcers arise in non-institutional
environments. 20
Pressure ulcers result in both increased length of hospital stay and hospital costs 21 and
increased nursing care time, as demonstrated by a study of long-term care patients. 22 Cellulitis,
osteomyelitis, and sepsis are morbid complications of untreated pressure ulcers. Increased
mortality has also been associated with pressure ulcers. 17
Opportunities for Impact
The passage of the Omnibus Budget Reconciliation Act of 1987 and subsequent
implementation regulations provided a written mandate for hospitalized and institutionalized
patients to receive regular assessment, preventive measures, and treatment of pressure ulcers.
There are no specific data on the current utilization of preventive measures by hospitals. The
pressure ulcer protocols often vary from institution to institution.
Study Designs
We identified a recent systematic review of pressure-relieving devices 23 that evaluated 37
randomized controlled trials (RCTs), 31 of which were focused on pressure ulcer prevention
(Table 27.1). Seven trials compared standard foam mattresses with various “low-technology”
supports. These low-technology supports were defined as beds with constant low-pressure
supports, including bead-, water-, or gel-filled supports; static air-filled supports; or foam or
Silicore-filled supports. Seven of the trials compared constant low-pressure devices with
alternating pressure devices. Six studies limited enrollment to orthopedic patients, 5 to patients
in intensive care units, and 3 to patients undergoing operations (ie, the study evaluated different
operating table surfaces). We identified no further RCTs of pressure ulcer prevention published
after the systematic review identified above.
Study Outcomes
All studies reported pressure ulcer development in both intervention and control groups.
Pressure ulcer grading systems were used in most trials. Typically, a 4-level grading system was
302

employed, ranging from Grade 1 (discolored skin) to Grade 4 (full-thickness skin lesions with
bone involvement).
Evidence for Effectiveness of the Practice
Many specialized beds appear to be effective in reducing the development of pressure
ulcers when compared with standard mattresses. For example, in 4 studies 24-27 cited by the
systematic review 23 that compared standard hospital foam mattresses to enhanced foam
alternatives, a summary relative risk of 0.29 (95% CI: 0.19-0.43) was calculated, favoring the
intervention group. Between-group comparisons of the previously defined low-technology
constant low-pressure devices did not yield clear conclusions. Similarly, in 7 RCTs the
comparison of alternating pressure devices with a variety of constant low-pressure devices (a
water mattress, foam pad, static air mattress, and foam overlays) showed no significant
difference in pressure ulcer development (summary RR 0.84, 95% CI: 0.57-1.23). 23 However, a
study of alternating pressure supports compared with standard foam mattresses did demonstrate
lower pressure ulcer development in the intervention group (RR 0.32; 95% CI 0.14-0.74). 28
Comparing pressure-reducing devices among themselves (versus against a standard mattress)
yields no significant differences in the prevention of pressure ulcers. These trials have been
summarized in a recent review. 29
In addition, 2 published trials evaluating different operating table-like surfaces suggest
reduction in pressure ulcer development with enhanced surfaces. 30, 31 In a well-designed RCT of
446 patients, Nixon and colleagues 31 showed that a dry gel-polymer pad placed on an operating
table decreased the incidence of new pressure ulcers by almost half—11% for intervention
patients vs. 20% for control patients placed on a standard operating table mattress (RR 0.46, 95%
CI: 0.26-0.82).
Several caveats temper the interpretation of studies of specialized pressure-relieving
surfaces. In general the studies had poor methodologic design, as the systematic review points
out. 23 The trials were mostly small, true baseline comparability was hard to confirm,
standardization of protocols was often unclear, and assessments were frequently unblinded.
Patient selection across trials was not consistent, and differences in pressure ulcer risk at
enrollment were difficult to compare across studies.
Potential for Harm
None reported.
Costs and Implementation
The costs of treating a pressure ulcer are estimated to range from $4000 to $40,000 for
newly developed ulcers. 32 Indeed, both hospital costs and length of stay are significantly higher
for patients who develop pressure ulcers during hospitalization, as noted earlier. 21 In the nursing
home in particular, failure to prevent this adverse outcome carries increasing liability—the
median settlement for pressure ulcer-related disputes was $250,000 between the years 1977 and
1987. 33 The cost of specialized beds and mattresses to prevent pressure ulcer development can be
high, ranging from $40 to $85 per day for low air-loss beds. 34 Specialized beds and intensive
nursing interventions all carry clear resource implications. Inman and colleagues 35 have
demonstrated the cost-effectiveness of an air suspension bed compared to a standard intensive
care unit bed. Yet cost-effectiveness studies of the many different pressure-relieving devices
have not been formally completed.
303

In terms of the feasibility of implementing these specific devices and following
guidelines for high-risk patients, both cost and time considerations must be examined. 36 Other
considerations relate to the design and functionality of a particular bed or mattress—for example
the ability of nursing staff to move and transfer patients placed on deeper or bulkier beds.
Finally, difficulty in accurately assessing changes in the incidence and prevalence of pressure
ulcers resulting from the institution of preventive measures is another barrier to documenting
progress. 37
Comment
Overall there is adequate evidence that specially designed surfaces effectively prevent the
development of pressure ulcers in high-risk patients, but the definition of high risk varies across
studies. The variety of pressure-relieving devices makes it difficult to recommend one over
another because there are few direct comparisons among the many different types of surfaces. Of
note, the treatment of established pressure ulcers is a separate topic, and the type of pressurerelieving
surface that is effective in treatment may not prove best for prevention. Appropriate
patient selection criteria need further development and refinement because the cost of many
prevention interventions is high. The necessity for larger RCTs to assess both clinical and costeffectiveness
of these specially designed mattresses is clear. Better descriptions of what
constitutes a “standard” bed or mattress, and improved reporting of baseline comparability
between experimental and control groups are also necessary to adequately interpret existing
studies. To better track progress in prevention, standardized strategies should be developed so
that accurate prevalence estimates can be documented.
Table 27.1. Studies of pressure ulcer prevention*
Participants and Setting Study Design,
Outcomes
Systematic review of 31 RCTs from
the US, UK and elsewhere assessing
pressure relieving interventions for
prevention of pressure ulcers 23
Level 1A,
Level 1
304
Relative Risk of Pressure Ulcer
Enhanced alternative foam mattresses
vs. standard hospital mattresses:
RR 0.29 (95% CI: 0.19-0.43)
Alternating pressure vs. constant low
pressure devices: RR 0.84
(95% CI: 0.57-1.23)
* CI indicates confidence interval; RCT, randomized controlled trial; and RR, relative risk.
References
1. Berlowitz DR, Wilking SV. Risk factors for pressure sores. A comparison of crosssectional
and cohort-derived data. J Am Geriatr Soc. 1989;37:1043-1050.
2. Brandeis GH, Ooi WL, Hossain M, Morris JN, Lipsitz LA. A longitudinal study of risk
factors associated with the formation of pressure ulcers in nursing homes. J Am Geriatr
Soc. 1994;42:388-393.
3. Allman RM, Goode PS, Patrick MM, Burst N, Bartolucci AA. Pressure ulcer risk factors
among hospitalized patients with activity limitation. JAMA. 1995;273:865-870.
4. Long term care facility resident assessment (RAI) user's manual. Version 2.0. Baltimore,
MD: Health Care Financing Administration; 1995.

5. Norton D, McLaren R, Exton-Smith AN. An investigation of geriatric nursing problems in
hospital. London: Churchill Livingstone; 1962.
6. Bergstrom N, Braden BJ, Laguzza A, Holman V. The Braden Scale for Predicting Pressure
Sore Risk. Nurs Res. 1987;36:205-210.
7. Xakellis GC, Frantz RA, Arteaga M, Nguyen M, Lewis A. A comparison of patient risk for
pressure ulcer development with nursing use of preventive interventions. J Am Geriatr Soc.
1992;40:1250-1254.
8. Agency for Health Care Policy and Research. Pressure Ulcer in Adults: Prediction and
Prevention. Clinical Practice Guideline, Number 3. Rockville, MD: Public Health Service,
U.S. Department of Health and Human Services; 1992.
9. Blom MF. Dramatic decrease in decubitus ulcers. Geriatr Nurs. 1985;6:84-87.
10. Moody BL, Fanale JE, Thompson M, Vaillancourt D, Symonds G, Bonasoro C. Impact of
staff education on pressure sore development in elderly hospitalized patients. Arch Intern
Med. 1988;148:2241-2243.
11. van der Cammen TJ, O'Callaghan U, Whitefield M. Prevention of pressure sores. A
comparison of new and old pressure sore treatments. Br J Clin Pract 1987;41:1009-1011.
12. Declair V. The usefulness of topical application of essential fatty acids (EFA) to prevent
pressure ulcers. Ostomy Wound Management. 1997;43:48-52,54.
13. Krasner D. Patient support surfaces. Ostomy Wound Management. 1991;33:47-56.
14. Brandeis GH, Morris JN, Nash DJ, Lipsitz LA. The epidemiology and natural history of
pressure ulcers in elderly nursing home residents. JAMA. 1990;264:2905-2909.
15. Smith DM. Pressure ulcers in the nursing home. Ann Intern Med. 1995;123:433-442.
16. Berlowitz DR, Bezerra HQ, Brandeis GH, Kader B, Anderson JJ. Are we improving the
quality of nursing home care: the case of pressure ulcers. J Am Geriatr Soc. 2000;48:59-62.
17. Allman RM, Laprade CA, Noel LB, Walker JM, Moorer CA, Dear MR, et al. Pressure
sores among hospitalized patients. Ann Intern Med. 1986;105:337-342.
18. Meehan M. National pressure ulcer prevalence survey. Adv Wound Care. 1994;7:27-
30,34,36-8.
19. Eckman KL. The prevalence of dermal ulcers among persons in the U.S. who have died.
Decubitus. 1989;2:36-40.
20. Guralnik JM, Harris TB, White LR, Cornoni-Huntley JC. Occurrence and predictors of
pressure sores in the National Health and Nutrition Examination survey follow-up. J Am
Geriatr Soc. 1988;36:807-812.
21. Allman RM, Goode PS, Burst N, Bartolucci AA, Thomas DR. Pressure ulcers, hospital
complications, and disease severity: impact on hospital costs and length of stay. Adv
Wound Care. 1999;12:22-30.
22. D'Hoore W, Guisset AL, Tilquin C. Increased nursing time requirements due to pressure
sores in long-term-care residents in Quebec. Clin Perform Qual Health Care. 1997;5:189-
194.
23. Cullum N, Deeks J, Sheldon TA, Song F, Fletcher AW. Beds, mattresses and cushions for
pressure sore prevention and treatment. In: The Cochrane Library, Issue 4, 2000. Oxford:
Update Software.
24. Hofman A, Geelkerken RH, Wille J, Hamming JJ, Hermans J, Breslau PJ. Pressure sores
and pressure-decreasing mattresses: controlled clinical trial. Lancet. 1994;343:568-571.
25. Santy JE, Butler MK, Whyman JD. A comparison study of 6 types of hospital mattress to
determine which most effectively reduces the incidence of pressure sores in elderly patients
305

with hip fractures in a District General Hospital. Report to Northern & Yorkshire Regional
Health Authority 1994.
26. Gray DG, Campbell M. A randomized clinical trial of two types of foam mattresses. J
Tissue Viability. 1994;4:128-132.
27. Collier ME. Pressure-reducing mattresses. J Wound Care. 1996;5:207-211.
28. Andersen KE, Jensen O, Kvorning SA, Bach E. Decubitus prophylaxis: a prospective trial
on the efficiency of alternating-pressure air-mattresses and water-mattresses. Acta Derm
Venereol (Stockholm). 1982;63:227-230.
29. Thomas DR. Issues and dilemmas in the prevention and treatment of pressure ulcers: a
review. J Gerontol. 2001;56A:M328-M40.
30. Aronovitch SA. A comparative, randomized, controlled study to determine safety and
efficacy of preventive pressure ulcer systems: preliminary analysis. Adv Wound Care.
1998;11:15-16.
31. Nixon J, McElvenny D, Mason S, Brown J, Bond S. A sequential randomised controlled
trial comparing a dry visco-elastic polymer pad and standard operating table mattress in the
prevention of post-operative pressure sores. Int J Nurs Stud. 1998;35:193-203.
32. Hibbs P. The economics of pressure ulcer prevention. Decubitus. 1988;1:32-38.
33. Bennett RG, O'Sullivan J, DeVito EM, Remsburg R. The increasing medical malpractice
risk related to pressure ulcers in the United States. J Am Geriatr Soc. 2000;48:73-81.
34. Ferrell BA, Keeler E, Siu AL, Ahn S, Osterweil D. Cost-effectiveness of low-air-loss beds
for treatment of pressure ulcers. J Gerontol. 1995;50A:M141-M46.
35. Inman KJ, Sibbald WJ, Rutledge FS, Clark BJ. Clinical utility and cost-effectiveness of an
air suspension bed in the prevention of pressure ulcers. JAMA. 1993;269:1139-1143.
36. Hu TW, Stotts NA, Fogarty TE, Bergstrom N. Cost analysis for guideline implementation
in prevention and early treatment of pressure ulcers. Decubitus. 1993;6:42-46.
37. Smith DM, Winsemius DK, Besdine RW. Pressure sores in the elderly: can this outcome
be improved? J Gen Intern Med. 1991;6:81-93.
306

Chapter 28. Prevention of Delirium in Older Hospitalized Patients
Joseph V. Agostini, MD
Dorothy I. Baker, PhD, RNCS
Sharon K. Inouye, MD, MPH
Sidney T. Bogardus, Jr., MD
Yale University Schools of Medicine and Public Health
Background
Delirium, or acute confusional state, is a common complication among hospitalized older
patients. Delirium is characterized by a sudden onset and fluctuating course, inattention, altered
level of consciousness, disorganized thought and speech, disorientation, and often behavioral
disturbance. As with other common geriatric syndromes, the etiology of delirium is
multifactorial. Previous research has identified a broad range of predisposing and precipitating
factors. 1-4 These include older age, cognitive or sensory impairments, dehydration, specific
medication usage (eg, psychoactive drugs), concurrent medical illness, and sleep deprivation.
The multifactorial nature of delirium suggests that intervention strategies targeting multiple
known risk factors might be effective in preventing its occurrence in hospitalized older patients.
In this chapter, we review multicomponent prevention programs that can be applied to a general
hospitalized patient population, not restricted to one admitting diagnosis (in keeping with the
crosscutting patient safety focus of the Report; Chapter 1). For example, a study comparing the
effect of postoperative analgesia using intravenous versus epidural infusions after bilateral knee
replacement surgery was not included. 5
Practice Description
A number of individual interventions have been used in efforts to prevent delirium. Some
could be considered part of general nursing practice, whereas others involve medical
assessments by physicians or consultants. General strategies to prevent delirium include use of
patient reorientation techniques (such as verbal reassurance, re-introduction of team members,
review of the daily hospital routine and patient schedule), environmental modifications (visible
clocks and calendars), and scheduled patient mobility. The number and complexity of these
interventions can vary, with individual nursing discretion usually determining how and when
these interventions are implemented. Patient education, 6 nursing staff education, 7 and family
involvement 8 are also useful. Approaches for primary prevention that incorporate physician
consultants or geriatric consultative teams 9-11 are reviewed elsewhere in this Report (see
Chapters 29 and 30).
Formal prevention programs target defined risk factors by implementing multiple
practices according to standardized protocols. For example, a recently reported, multicomponent
strategy focused on 6 risk factors and successfully developed intervention protocols to address
each of them. 12 Patients with cognitive impairment received daily orientation interventions and
3-times daily cognitive stimulation activities. To target sleep impairment, patients received nonpharmacologic
sleeping aids (eg, back massage and relaxation tapes), while hospital staff
engaged in noise-reduction strategies such as setting beepers to vibrate and using silent pill
crushers. Immobility was addressed with a 3-times daily exercise protocol adapted for use with
bed-bound and ambulatory patients. Sensory impairments were addressed by providing devices
such as auditory amplifiers, visual aids, and larger size push-button phones. Patients with
307

evidence of dehydration received standardized repletion interventions. A geriatric nurse
specialist and staff assisted by trained volunteers carried out all the interventions.
Prevalence and Severity of the Target Safety Problem
The target safety problem is the primary prevention of delirium, rather than the
treatment 13 of existing delirium. In the United States, delirium affects an estimated 2.3 million
hospitalized elders annually, accounting for 17.5 million inpatient days, and leading to more than
$4 billion in Medicare costs (1994 dollars). 12 Studies have found that delirium in hospitalized
patients contributes to longer lengths of stay, 14 increased mortality, 15-17 and increased rates of
institutional placement. 18, 19 New cases of delirium occur in approximately 15% to 60% of
hospitalized older patients, depending on the number of risk factors present at
admission. 4,15,18,20,21 Moreover, because many cases of delirium go unrecognized during
hospitalization and because symptoms may persist for months after discharge, 22 these may be
conservative estimates. Safety practices to reduce delirium may thus have substantial impact on
the health and well-being of older patients in hospitals. These practices may also impact nursing
home residents and other institutionalized patients, but our practice review did not identify any
studies carried out among these patient populations.
Opportunities for Impact
It is difficult to estimate the extent of existing practices aimed at decreasing delirium. A
comprehensive model, the Hospital Elder Life Program, 23 which incorporates the delirium
interventions reviewed in one study in this chapter, 12 is presently in the initial dissemination
phase at 6 replication sites, with 16 hospitals on a waiting list. Present evidence suggests that few
facilities currently have intervention programs designed for the primary prevention of delirium.
The opportunity for impact in nursing homes and other long-term care facilities is great, but thus
far studies have not targeted these settings.
Study Designs
Cole 24 conducted a structured search of the medical literature and identified 10
intervention trials to prevent delirium in hospitalized patients. Of these, we excluded one study
of much younger patients (mean age, 49 years) 25 and one study that incorporated interventions
not applicable to most hospitalized elders (eg, early surgery, prevention and treatment of perioperative
blood pressure falls). 26 Three used psychiatric consultations 27-29 which did not fit our
criteria for risk factor intervention (see Chapter 29 for similar studies). Table 28.1 lists the
remaining 5 studies 6, 8, 30-32 and a later study, 12 which is the largest controlled trial to date.
Study Outcomes
All of the studies in Table 28.1 reported delirium or confusion symptoms as an outcome
measure. Each study, however, used a different instrument to identify delirium: DSM-III, 33 the
Confusion Assessment Method, 34 the Short Portable Mental Status Questionnaire, 35 or a scoring
system based on delirium symptoms.
Evidence for Effectiveness of the Practice
The earliest studies, by Owens 6 and Chatham, 8 focused on the effects of patient and
family education, respectively. Delirium symptoms modestly improved but achieved statistical
significance in only 5 of the 11 symptom categories reported in the latter study. Both studies
were limited by small numbers of patients, non-standardized interventions, and minimal data on
308

aseline co-morbidities of the enrolled patients. The study by Williams and colleagues, 32 which
targeted a population at high risk for delirium (older patients with hip fracture), also
demonstrated a statistically significant reduction in delirium symptoms by targeting
environmental nursing interventions and patient education. Two subsequent studies did not show
a reduction in delirium. The low incidence of delirium (only 3 cases in 30 intervention patients)
in the study by Nagley et al 30 created inadequate power to detect a significant effect with only 60
total patients. Although a high percentage of patients experienced delirium in the study by
Wanich et al, 31 79% of cases were diagnosed at the time of admission (prevalent rather than
incident cases) and therefore could not have been prevented by the intervention. Both of these
studies may also have suffered from contamination bias. The greatest benefit in delirium
prevention, a 40% risk reduction, occurred in the study by Inouye et al, 12 a carefully designed
and implemented hospital program targeting 6 well-recognized risk factors for delirium, in
which adherence to each intervention protocol was tracked. The intervention reduced the number
and severity of patients’ risk factors and was successful in preventing patients’ first delirium
episode.
Potential for Harm
None noted.
Costs of Implementation
The only recent estimate of cost per case of delirium prevented was $6341 in a delirium
prevention trial, 12 which is less than the cost associated with prevention of other hospital
complications such as falls. A further analysis of the same patients reveals that the
multicomponent strategy is cost-effective for those at intermediate risk of delirium, but not for
those at highest risk. 36
Comment
The literature for delirium prevention studies is small, and the methodologic quality of
many studies is poor. However, one high quality study 12 has demonstrated that multicomponent
interventions can prevent incident delirium in hospitalized patients. The interventions have high
face validity and are both feasible and transportable across institutions and hospital units,
suggesting that implementation in different practice settings would be practical. Implementing a
multicomponent intervention on a hospital-wide basis throughout the United States would
require significant commitment from hospital staff. Programs such as the Hospital Elder Life
Program 23 can be readily integrated into hospital practice and have been successful in preventing
both cognitive and functional decline using targeted, practical interventions. Others of these
practices could be incorporated by either support staff or trained volunteers, which may save
resources and underscore the fact that many common sense interventions do not require a larger
professional staff. Future studies should focus on refining the most effective multifactorial
programs, determining the optimal combination of interventions, defining appropriate target
populations based on delirium risk, demonstrating effectiveness across multiple clinical sites,
and disseminating the most cost-effective practices.
309

Table 28.1. Six studies of delirium prevention*
Study Study Setting Interventions Study Design
Outcomes
Chatham,
1978 8
Inouye,
1999 12
Nagley,
1986 30
Owens,
1982 6
Wanich,
1992 31
Williams,
1985 32
20 surgical
patients in a
university
affiliated
hospital, 1977
852 patients in a
university
hospital, 1995-
1998
60 patients at a
university
affiliated
hospital
64 surgical
patients in a
university
hospital
235 patients in a
university
hospital, 1986-
1987
227 orthopedic
patients in 4
hospitals
• Family education
• Patient education
Targeted 6 risk factors:
• Cognitive impairment
• Immobility
• Visual impairment
• Hearing impairment
• Dehydration
• Sleep deprivation
16 interventions, including:
• Orientation strategies
• Providing sensory aides
• Ambulation
• Hydration measures
• Nursing interaction
310
Level 2,
Level 1
Level 2,
Level 1
Level 2,
Level 1
• Patient education Level 2,
Level 1
• Nursing education
• Caregiver education
• Orientation strategies
• Mobilization
• Environmental
modifications
• Medication evaluation
• Patient education
• Orientation strategies
• Providing sensory aides
Level 2,
Level 1
Level 2,
Level 1
* CI indicates confidence interval; OR, odds ratio.
† Delirium rate is the percentage of patients with one or more episodes of delirium.
Results†
Delirium symptoms rate:
intervention resulted in
improvement in 5 of 11
areas—orientation,
appropriateness, confusion,
delusions, and sleep (p0.05)
Delirium symptoms rate:
intervention 59%, control
78% (p>0.05)
Delirium rate: intervention
19%, control 22% (p=0.61)
Delirium symptoms rate:
intervention 43.9%, control
51.5% (p

References
1. Elie M, Cole MG, Primeau FJ, Bellavance F. Delirium risk factors in elderly hospitalized
patients. J Gen Intern Med. 1998;13:204-212.
2. Inouye SK, Viscoli CM, Horwitz RI, Hurst LD, Tinetti ME. A predictive model for
delirium in hospitalized elderly medical patients based on admission characteristics. Ann
Intern Med. 1993;119:474-481.
3. Schor JD, Levkoff SE, Lipsitz LA, Reilly CH, Cleary PD, Rowe JW, et al. Risk factors for
delirium in hospitalized elderly. JAMA. 1992;267:827-831.
4. Inouye SK, Charpentier PA. Precipitating factors for delirium in hospitalized elderly
persons. Predictive model and interrelationship with baseline vulnerability. JAMA.
1996;275:852-857.
5. Williams-Russo P, Urquhart BL, Sharrock NE, Charlson ME. Post-operative delirium:
predictors and prognosis in elderly orthopedic patients. J Am Geriatr Soc. 1992;40:759-
767.
6. Owens JF, Hutelmyer CM. The effect of preoperative intervention on delirium in cardiac
surgical patients. Nurs Res. 1982;31:60-62.
7. Rapp CG, Onega LL, Tripp-Reimer T, Mobily P, Wakefield B, Kundrat M, et al. Unitbased
acute confusion resource nurse: an educational program to train staff nurses.
Gerontologist. 1998;38:628-632.
8. Chatham MA. The effect of family involvement on patients' manifestations of
postcardiotomy psychosis. Heart Lung. 1978;7:995-999.
9. Hogan DB, Fox RA, Gadley BW, Mann OE. Effect of a geriatric consultation service on
management of patients in an acute care hospital. CMAJ. 1987;136:713-717.
10. Cole MG, Fenton FR, Englesmann F, Mansouri I. Effectiveness of geriatric psychiatry
consultation in an acute care hospital: a randomized clinical trial. J Am Geriatr Soc.
1991;39:1183-1188.
11. Marcantonio ER, Flacker JM, Wright RJ, Resnick NM. Reducing delirium after hip
fracture: a randomized trial. J Am Geriatr Soc. 2001;49:In press.
12. Inouye SK, Bogardus ST, Jr., Charpentier PA, Leo-Summers L, Acampora D, Holford TR,
et al. A multicomponent intervention to prevent delirium in hospitalized older patients. N
Engl J Med. 1999;340:669-676.
13. Meagher DJ. Delirium: optimising management. BMJ. 2001; 322:144-149.
14. Pompei P, Foreman M, Rudberg MA, Inouye SK, Braund V, Cassel CK. Delirium in
hospitalized older persons: outcomes and predictors. J Am Geriatr Soc. 1994;42:809-815.
15. Cameron DJ, Thomas RI, Mulvihill M, Bronheim H. Delirium: a test of the Diagnostic and
Statistical Manual III criteria on medical inpatients. J Am Geriatr Soc. 1987;35:1007-1010.
16. Weddington WW. The mortality of delirium: an underappreciated problem?
Psychosomatics. 1982;23:1232-1235.
17. Cole MG, Primeau FJ. Prognosis of delirium in elderly hospital patients. CMAJ.
1993;149:41-46.
18. Francis J, Martin D, Kapoor WN. A prospective study of delirium in hospitalized elderly.
JAMA. 1990;263:1097-1101.
19. Inouye SK, Rushing JT, Foreman MD, Palmer RM, Pompei P. Does delirium contribute to
poor hospital outcomes? A three-site epidemiologic study. J Gen Intern Med. 1998;13:234-
242.
311

20. Foreman MD. Confusion in the hospitalized elderly: incidence, onset, and associated
factors. Res Nurs Health. 1989;12:21-29.
21. Rockwood K. Acute confusion in elderly medical patients. J Am Geriatr Soc. 1989;37:150-
154.
22. Levkoff SE, Evans DA, Liptzin B, Cleary PD, Lipsitz LA, Wetle TT, et al. Delirium. The
occurrence and persistence of symptoms among elderly hospitalized patients. Arch Intern
Med. 1992;152:334-340.
23. Inouye SK, Bogardus ST, Jr., Baker DI, Leo-Summers L, Cooney LM, Jr. The Hospital
Elder Life Program: a model of care to prevent cognitive and functional decline in older
hospitalized patients. J Am Geriatr Soc. 2000;48:1697-1706.
24. Cole MG, Primeau F, McCusker J. Effectiveness of interventions to prevent delirium in
hospitalized patients: a systematic review. CMAJ. 1996;155:1263-1268.
25. Budd S, Brown W. Effect of a reorientation technique on postcardiotomy delirium. Nurs
Res. 1974;23:341-348.
26. Gustafson Y, Brannstrom B, Berggren D, Ragnarsson JI, Sigaard J, Bucht G, et al. A
geriatric-anesthesiologic program to reduce acute confusional states in elderly patients
treated for femoral neck fractures. J Am Geriatr Soc. 1991;39:6556-62.
27. Layne OL, Jr., Yudofsky SC. Postoperative psychosis in cardiotomy patients. The role of
organic and psychiatric factors. N Engl J Med. 1971;284:518-520.
28. Schindler BA, Shook J, Schwartz GM. Beneficial effects of psychiatric intervention on
recovery after coronary artery bypass graft surgery. Gen Hosp Psychistry. 1989;11:358-
364.
29. Lazarus HR, Hagens JH. Prevention of psychosis following open-heart surgery. Am J
Psychiatry. 1968;124:1190-1195.
30. Nagley SJ. Predicting and preventing confusion in your patients. J Gerontol Nurs.
1986;12:27-31.
31. Wanich CK, Sullivan-Marx EM, Gottlieb GL, Johnson JC. Functional status outcomes of a
nursing intervention in hospitalized elderly. Image J Nurs Sch. 1992;24:201-207.
32. Williams MA, Campbell EB, Raynor WJ, Mlynarczyk SM, Ward SE. Reducing acute
confusional states in elderly patients with hip fractures. Res Nurs Health. 1985;8:329-337.
33. American Psychiatric Association. Diagnostic and statistical manual of mental disorders
(3rd edition). ed. Washington, DC: American Psychiatric Press; 1980.
34. Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI. Clarifying
confusion: the confusion assessment method. A new method for detection of delirium. Ann
Intern Med. 1990;113:941-948.
35. Pfeiffer E. A short portable mental status questionnaire for assessment of organic brain
deficit in elderly patients. J Am Geriatr Soc. 1979;23:433-438.
36. Rizzo JA, Bogardus ST, Jr., Leo-Summers L, Williams CS, Acampora D, Inouye SK. A
multicomponent targeted intervention to prevent delirium in hospitalized older patients:
what is the economic value? Medical Care. 2001;In press.
312

Chapter 29. Multidisciplinary Geriatric Consultation Services
Joseph V. Agostini, MD
Dorothy I. Baker, PhD, RNCS
Sharon K. Inouye, MD, MPH
Sidney T. Bogardus, Jr., MD
Yale University Schools of Medicine and Public Health
Background
Multidisciplinary geriatric consultation teams provide comprehensive assessment of
physical, emotional, and functional status in older persons and make recommendations regarding
prevention and management of common geriatric syndromes, functional impairments, and other
problems. Teams vary in composition but usually include a geriatrician, nurse, social worker and
other health professionals such as rehabilitative therapists, psychologists, and dieticians. Their
expertise may improve the safety of hospitalized elders (or nursing home residents) by reducing
hospital-acquired complications such as falls, delirium, functional decline, and “preventable”
deaths. Consultation teams and inpatient geriatric evaluation and management (GEM) units share
this multidisciplinary approach but differ in who controls the implementation of
recommendations. A patient’s primary physician decides whether to implement a consultation
team’s recommendations, whereas geriatric teams have direct control over implementation in
GEM units. As this difference may impact effectiveness, GEM units are reviewed separately
(Chapter 30). Multidisciplinary consultation services are available to elders living in the
community, however this chapter reviews the evidence for geriatric consultation improving the
safety of hospitalized patients or nursing home residents.
Practice Description
The structure of consultation teams, the types of evaluations they routinely perform, and
the recommendations they make vary from institution to institution. A review of published
studies reveals some common features. A team has at least one physician, a nurse practitioner or
similarly professionally-trained provider, rehabilitative experts, and usually a social worker.
Assessments typically include measures of mobility and functional status, mental status
examinations, psychological screening, evaluation of common geriatric problems (eg, risk for
falls, incontinence, and polypharmacy), and plans for rehabilitation and/or appropriate placement
at the time of discharge. The team notes its recommendations in the hospital chart,
communicates them to the physician directing care for the patient, and provides follow-up until
the patient is discharged.
Prevalence and Severity of the Target Safety Problem
Patients aged 65 years and older account for almost half of inpatient hospitalization
days. 1 In 1996, they comprised 13% of the population yet accounted for 38% of the
approximately 31 million discharges from non-government, acute care hospitals. 2 The actual
target population is smaller because, although all hospitalized elders are at risk for
complications, some patients are unlikely to benefit from multidisciplinary geriatric consultation.
Strategies to target geriatric services to those patients most likely to benefit have been reviewed
in the literature. 3 The characteristics associated with potential benefit include advanced age (eg,
over 75 years old), specific geriatric conditions (eg, falls or confusion), functional impairments
313

(eg, limitations in bathing, feeding, or transferring), and psychosocial impairments (depression
or living alone).
As a patient safety practice, consultations may decrease the occurrence of iatrogenic
complications such as functional decline related to hospitalization, delirium, and falls. Functional
decline occurs in 25 to 60% of older persons after entering acute care, due to the interaction of a
patient’s existing co-morbidities with the hospital environment. 4 It results in worsened cognitive
status and physical functioning due to the stressors of hospitalization (bedrest and immobility,
medical procedures and pharmacotherapy, and the hospital environment) in older patients. For
information on the prevalence and severity of falls and delirium, see Chapters 26 and 28,
respectively.
Mortality is another important clinical outcome that may be affected. The number of
deaths that might be prevented by implementation of this practice is unknown, although the most
common medical diagnoses of patients enrolled in these consultation studies (cardiac,
pulmonary, and gastrointestinal disorders) reflect the prevalent admitting medical diagnoses for
all older patients. The in-hospital, all-cause mortality in the reviewed studies (approximately 5 to
15%) provides a context against which one can consider the potential for improvement if these
practices influence mortality.
Opportunities for Impact
The number of hospitals with multidisciplinary geriatric consultation services is not
reported in the literature. However, data from the American Hospital Association 5 indicate that
fewer than half of hospitals offer comprehensive geriatric assessment.* Researchers in the field
believe that even in those hospitals with consultation services, only a minority of the patients
most likely to benefit are being referred. Thus, if the practice is effective, there is substantial
opportunity for improvement by increasing its utilization in this vulnerable patient population.
Study Designs
A structured literature search identified 14 controlled trials: 12 randomized, 6-17 1
alternate-allocation, 18 and 1 prospective cohort study 19 (see Table 29.1). Four of the 14 articles
report different outcome measures from the same clinical trial. 6-9 One study focuses on nursing
home residents; 14 all other studies were of hospitalized patients. Three of the studies were
performed in Canada and one in the United Kingdom. 16-19 Two trials were limited to elderly
patients with hip fractures. 15, 17 In the study by Fretwell et al, patients were admitted to a medical
ward designated for seniors and staffed with specially trained nurses. 13 Because the team still
functioned in a consultative role and could not implement its own recommendations, the study is
included here rather in the chapter on GEM units (Chapter 30).
* Of the 4953 acute medical/surgical hospitals in the American Hospital Association (AHA)
database, 4398 (89%) responded to the AHA 1999 Annual Survey. Of responding hospitals,
1823 (41%) indicated availability of “geriatric services,” which was defined as providing one or
more of the following: comprehensive geriatric assessment, adult day care, Alzheimer’s
diagnostic-assessment services, geriatric acute care units, and/or geriatric clinics. A
conservative, upper-limit estimate assuming all 555 non-responding hospitals have “geriatric
services” would be 48%. As the survey does not ask the availability of each type of geriatric
service, the percentage of hospitals offering inpatient comprehensive geriatric assessment based
on the AHA Survey data can only be described as “less than 48%” (how much less is unknown).
314

Study Outcomes
Ten studies reported functional status outcomes, measured by the Katz 20 or Lawton 21
index of activities of daily living, or the Barthel Index. 22 Marcantonio et al 15 measured the
occurrence and severity of delirium defined according to the Confusion Assessment Method
criteria. 23 Ray and colleagues reported the proportion of recurrent fallers and the rate of injurious
falls in nursing home patients during one year. 14 Becker et al measured 6 classes of hospitalacquired
complications: medication-related, procedures, infections, trauma or injury (eg, falls
and pressure sores), psychiatric, and other (eg, urinary retention, fecal impaction). 7 Eight trials
reported all-cause mortality in-hospital, at 6 months, or at one year. 8,10-13,16,18,19 A recent metaanalysis
24 incorporated unpublished mortality data from several other studies 6,17 reviewed here.
Other clinically relevant outcomes were changes in pharmacotherapy prescribed, length of
hospital stay, and discharge location.
Evidence for Effectiveness of the Practice
Two non-blinded trials showed a statistically significant improvement in patients’
functional ability. 17,18 Kennie et al targeted a population at high risk for functional decline during
hospitalization: elderly women with hip fractures. 17 In the study by Hogan et al, 18 the difference
was significant at one year but not at 3 or 6 months, suggesting that the intervention group’s
post-discharge follow-up by a geriatric team may have accounted for the difference rather than
prevention of iatrogenic functional decline in-hospital. The study by Thomas and colleagues 12
showed a trend towards improved functional status. No other study reported improved functional
outcomes.
The trial by Becker et al 7 showed no significant difference in the incidence of hospitalacquired
complications between intervention and control groups. Two studies that targeted
specific high-risk populations did show benefit. 14,15 In the study by Marcantonio et al, a
multidisciplinary consultation including assessment and targeted recommendations per a
structured protocol in 10 domains (including pain treatment, bowel/bladder function, nutrition,
pain treatment, mobility, and environmental stimuli) resulted in a significant decrease in
perioperative delirium in patients with hip fracture. 15 Ray et al enrolled nursing home residents
65+ years of age who had fallen in the previous year and had a possible problem in at least one
of 4 safety domains: environmental safety, wheelchair use, psychotropic drug use, or mobility. 14
Patients who received care from the consultation team, including structured assessments and
specific recommendations in these safety domains, experienced a significant reduction in the rate
of recurrent falls (43.8% intervention group vs. 54.1% control group, p=0.03). 14
The reported randomized clinical trials yielded mixed results for the outcome of all-cause
mortality, with most studies demonstrating no benefit. The study by Thomas 12 reported a
statistically significant improvement in mortality at 6 months, but Gayton et al 19 reported only a
trend toward improvement at 6 months. Neither of Hogan’s studies found in-hospital mortality
benefits. One study 16 showed improved mortality at 4 months and one 18 at 6 months but these
benefits were not sustained at one year. Hospital-acquired complications would be expected to
reduce in-hospital or short-term mortality, so the survival benefit observed many months after
hospitalization in these studies suggests that other carry-over effects (eg, improved medication
regimens) or better post-discharge care may be influencing these results. According to a metaanalysis
24 the summary odds ratio for 6-month mortality in 8 of the studies cited 6,10,12,13,16-19 was
0.77 (95% CI: 0.62-0.96), but the effect on 12-month mortality was not statistically significant.
The authors tested for heterogeneity of outcomes before pooling results of the trials (p=0.07). Of
315

note, the large trial (n=2353) by Reuben et al 11 was not eligible for inclusion in the meta-analysis
because it was published later. Because it was larger than all other studies combined, its effect on
the pooled estimate of 6-month mortality would be to reduce any statistically significant
differences between intervention and study groups, since no survival advantage was reported at
up to one year in the study (p=0.89 for survival curve).
Potential for Harm
No harm attributable to the geriatric consultation was reported in the trials.
Costs and Implementation
Implementation of the multidisciplinary team entails logistic planning to determine the
number and type of consultative team members, and human resource coordination regarding
time allocation and staffing. Few studies included data on costs of the practice, such as hospital
costs incurred by assembly of the consultation team. Fretwell and colleagues, 13 however, have
reported hospital charges in their study of 436 patients in a university-affiliated hospital. The
percentage of patients exceeding DRG reimbursement for hospitalization was similar in both
intervention and control groups, 69.7% and 71.2%, respectively. Winograd 25 reported that the
cost of screening about 1200 patients to identify suitable candidates for consultation (using
predefined criteria discussed in the paper) could be accomplished by employing a trained
employee working one-quarter time, at a cost (in 1998 dollars) of about $7000 over the course of
one year.
Comment
Inpatient geriatric consultation may have an impact on care for the hospitalized older
patient, but the potential improvement in patient-safety outcomes is unclear. All-cause mortality
differences may be due to differences in patient selection, and the data for improvement in
functional outcomes suggests that certain patients may experience greater benefit than others.
Appropriate targeting of services to patients at high risk for adverse outcomes such as falls and
delirium seems to result in benefit. Consequently, multidisciplinary geriatric consultation and
other efforts directed towards preventing iatrogenic functional decline, the most common
complication of older hospitalized patients, deserve careful attention.
Identified problems in the reviewed studies include inadequate targeting of individuals
who would most benefit from the intervention, potential between-group cross-contamination,
and differences in local expertise in carrying out recommended interventions. Lack of
effectiveness in some studies may reflect poor compliance with team suggestions or inadequate
staffing to implement a consultant’s recommendations, regardless of desire to comply. Lack of
control over the direct management of the patient could also represent a serious shortcoming that
limits effectiveness of this practice.
Multidisciplinary geriatric consultative teams, in contrast to specialized geriatric
evaluation and management (GEM) units, provide expertise in geriatric care throughout a
hospital, but in a consultative role. In comparing this strategy with GEM units or Acute Care for
Elders (ACE) units, several differences should be noted. Multidisciplinary teams are less
expensive to organize and can be implemented within a shorter period of time. Since older
patients reside throughout an institution, there is also greater opportunity to reach a larger
number of patients when the consultation team is not single-unit based. There is no bed limit,
and the capacity of the team to provide interventions is therefore limited by their available time
rather than the number of beds in any one unit. The resources required to assemble an
316

experienced geriatric team in a hospital that has no pre-existing geriatric expertise remains an
important consideration. In addition, costs associated with enhancing and monitoring adherence
with recommendations should be included when designing an effective program. Physical
redesign of the unit environments to accommodate special needs (eg, special flooring, bed
layout, reorienting devices) is likewise not part of this practice. Specially trained geriatric nurses
are also not available equally throughout the hospital, in contrast to designated geriatric inpatient
units, which have nurses focused exclusively on care of the older patient.
Notwithstanding these considerations, the practice of multidisciplinary geriatric
consultation services has high face validity. More research is needed to evaluate which patients
might receive maximal benefit for the associated resource commitment. Other areas for further
research include examining the problems most appropriate for geriatric assessment and
consultation in the hospital, developing strategies to improve adherence to and execution of
recommendations, and identifying the components of a successful and cost-effective consultation
team.
317

Table 29.1. Studies of multidisciplinary geriatric consultation services
Study Setting and
Participants
Allen,
1998 6
Becker,
1987 7
Saltz, 1988 8
McVey,
1989 9
Fretwell,
1990 13
Gayton,
1982 19
Hogan,
1987 16
Hogan,
1990 18
Kennie,
1988 17
185 patients at a VA
hospital, 1983-1984
436 patients at a
university-affiliated
hospital, 1985-1987
222 patients at a
Canadian universityaffiliated
hospital, 1982-
1984
113 patients at a
Canadian tertiary care
hospital, 1984
132 patients at a
Canadian hospital, 1985
144 orthopedic patients
at a U.K. district
hospital, year not stated
Study
Design,
Outcomes
Level 1,
Level 1
Level 1,
Level 1
Level 2,
Level 1
Level 1,
Level 1
Level 1,
Level 1
Level 1,
Level 1
318
Results
No significant differences in hospitalacquired
complications (overall 38%
for both groups)
No statistically significant improvement
in functional status (activities of daily
living)
No statistically significant differences in
rehospitalization or placement
Compliance with recommendations:
71.7% overall (from 47-95% for
selected interventions)
No significant difference in mortality at
discharge
No significant differences in length of
stay, physical or cognitive function, or
hospital charges
No significant mortality difference up to 6
months follow-up, but trend favoring
intervention group
No significant differences in functional
status, length of stay, or mental status
between study groups
Mortality at 4 months lower in the
intervention group (p

Marcantoni
o, 2001 15
Ray,
1997 14
Reuben,
1995 11
Thomas,
1993 12
Winograd,
1993 10
126 orthopedic patients
at an academic medical
center, year not stated
482 residents in 14
nursing homes, 1993-
1995
2353 patients at 4 HMOrun
hospitals, 1991-1994
120 patients at a
community hospital, year
not stated
197 men at a VA
hospital, 1985-1989
Level 1,
Level 1
Level 1,
Level 1
Level 1,
Level 1
Level 1,
Level 1
Level 1,
Level 1
319
Occurrence of delirium: 32% vs. 50% in
control group (p=0.04)
Adherence to recommendations: 77%
Lower rate of recurrent falls: 19% vs.
54% in control group (p=0.03)
Trend toward lower mean rate of injurious
falls
No statistically significant differences in
mortality at up to one-year follow-up
No significant change in functional status
at 3 or 12 months
Reduced 6-month mortality: 6% vs. 21%
controls (p=0.01)
Trend toward improved functional status
in the intervention group
Hospital readmission in 6-months
significantly lower in the intervention
group
No significant mortality differences
between groups
No significant change in physical
function, length of stay, or placement
between groups
Compliance with all recommendations:
67%

References
1. A profile of older Americans: 2000. [Washington, DC]: Administration on Aging. U.S.
Department of Health and Human Services, 2000.
2. Kramarow E, Lentzner E, Rooks R, Weeks J, Saydah S. Health and aging chartbook.
Health, United States, 1999. Hyattsville, Maryland: National Center for Health Statistics,
1999.
3. Winograd CH. Targeting strategies: an overview of criteria and outcomes. J Am Geriatr
Soc. 1991;39S:25S-35S.
4. Palmer RM. Acute hospital care of the elderly: minimizing the risk of functional decline.
Cleve Clin J Med. 1995;62:117-128.
5. American Hospital Association. Health Statistics, 2001 edition. Chicago, Illinois: Health
Forum, 2001.
6. Allen CM, Becker PM, McVey LJ, Saltz C, Feussner JR, Cohen HJ. A randomized,
controlled clinical trial of a geriatric consultation team. Compliance with
recommendations. JAMA. 1986;255:2617-2621.
7. Becker PM, McVey LJ, Saltz CC, Feussner JR, Cohen HJ. Hospital-acquired complications
in a randomized controlled clinical trial of a geriatric consultation team. JAMA.
1987;257:2313-2317.
8. Saltz CC, McVey LJ, Becker PM, Feussner JR, Cohen HJ. Impact of a geriatric
consultation team on discharge placement and repeat hospitalization. Gerontologist.
1988;28:344-350.
9. McVey LJ, Becker PM, Saltz CC, Feussner JR, Cohen HJ. Effect of a geriatric consultation
team on functional status of elderly hospitalized patients. Ann Intern Med. 1989;110:79-84.
10. Winograd CH, Gerety MB, Lai NA. A negative trial of inpatient geriatric consultation.
Lessons learned and recommendations for future research. Arch Intern
Med.1993;153:2017-2023.
11. Reuben DB, Borok GM, Wolde-Tsadik G, et al. A randomized trial of comprehensive
geriatric assessment in the care of hospitalized patients. N Engl J Med. 1995;332:1345-
1350.
12. Thomas DR, Brahan R, Haywood BP. Inpatient community-based geriatric assessment
reduces subsequent mortality. J Am Geriatr Soc. 1993;41:101-104.
13. Fretwell MD, Raymond PM, McGarvey ST, et al. The Senior Care Study. A controlled trial
of a consultative/unit-based geriatric assessment program in acute care. J Am Geriatr Soc.
1990;38:1073-1081.
14. Ray WA, Taylor JA, Meador KG, et al. A randomized trial of a consultation service to
reduce falls in nursing homes. JAMA. 1997;278:557-562.
15. Marcantonio ER, Flacker JM, Wright RJ, Resnick NM. Reducing delirium after hip
fracture: a randomized trial. J Am Geriatr Soc. 2001;49:In press.
16. Hogan DB, Fox RA, Badley BW, Mann OE. Effect of a geriatric consultation service on
management of patients in an acute care hospital. CMAJ. 1987;136:713-717.
17. Kennie DC, Ried J, Richardson IR, Kiamari AA, Kelt C. Effectiveness of geriatric
rehabilitative care after fractures of the proximal femur in elderly women: a randomised
clinical trial. BMJ. 1988;297:1083-1086.
18. Hogan DB, Fox RA. A prospective controlled trial of a geriatric consultation team in an
acute-care hospital. Age Ageing.1990;19:107-113.
320

19. Gayton D, Wood-Dauphinee S, de Lorimer M, Tousignant P, Hanley J. Trial of a geriatric
consultation team in an acute care hospital. J Am Geriatr Soc. 1987;35:726-736.
20. Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of illness in the aged.
The index of ADL: a standardized measure of biological and psychosocial function. JAMA.
1963;185:914-919.
21. Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental
activities of daily living. Gerontologist. 1969;9:179-186.
22. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J.
1965;14:61-65.
23. Inouye SK, Van Dycke CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI. Clarifying
confusion: the Confusion Assessment Method. Ann Intern Med. 1990;113:941-948.
24. Stuck AE, Siu AL, Wieland GD, Adams J, Rubenstein LZ. Comprehensive geriatric
assessment: a meta-analysis of controlled trials. Lancet. 1993;342:1032-1036.
25. Winograd CH, Gerety MB, Brown E, Kolodny V. Targeting the hospitalized elderly for
geriatric consultation. J Am Geriatr Soc. 1988;36:1113-1119.
321

322

Chapter 30. Geriatric Evaluation and Management Units for Hospitalized
Patients
Joseph V. Agostini, MD
Dorothy I. Baker, PhD, RNCS
Sidney T. Bogardus, Jr., MD
Yale University Schools of Medicine and Public Health
Background
Inpatient care of elders with multiple co-morbidities requires close attention to the special
needs and geriatric syndromes that arise in this vulnerable population. 1,2 One strategy to address
the risks of hospitalization is to provide care by a multidisciplinary team in a dedicated geriatric
unit, principally Geriatric Evaluation and Management (GEM) Units. This model of care, in
many ways similar to coordinated stroke units, 3 may improve mortality and other clinically
relevant outcomes compared with outcomes achieved on a general medical ward. An alternative
strategy, reviewed elsewhere in this Report, is the comprehensive geriatric consultation service,
analogous to a typical medical consultation team (see Chapter 29).
In a GEM unit, a multidisciplinary team provides comprehensive geriatric assessment,
detailed treatment plans, and attention to the rehabilitative needs of older patients. A typical team
is composed of a geriatrician, clinical nurse specialist, social worker, and specialists from such
fields as occupational and physical therapy, nutrition, pharmacy, audiology, and psychology.
GEM units are typically separate hospital wards that have been redesigned to facilitate care of
the geriatric patient. Multidisciplinary team rounds and patient-centered team conferences are
hallmarks of care on these units, which, in contrast to geriatric consultation services, have direct
control over the implementation of team recommendations.
Practice Description
In all the studies reviewed in this chapter, the GEM unit team included a physician
experienced in geriatric medicine, skilled geriatric nurses and rehabilitation specialists. (The
latter may not have been on site but were accessible). The teams completed multidimensional
geriatric assessments, conducted interdisciplinary team rounds, and provided comprehensive
discharge planning. Units were physically separate wards that were designed to facilitate
geriatric care. Acute Care of Elders (ACE) Units 4 incorporate the GEM unit design with
additional enhancements and admit patients with acute illnesses. ACE units often provide
improved flooring and lighting, reorienting devices, and other environmental improvements such
as common rooms for patient use. For example, in the ACE unit studied by Landefeld and
colleagues 5 the GEM unit concept was enhanced by the use of more nurse-initiated protocols and
a greater number of environmental and design modifications.
Both styles of unit emphasize the early assessment of risk factors for iatrogenic
complications and the prevention of functional decline. 4,6 Studies of both are included in this
chapter.
323

Prevalence and Severity of the Target Safety Problem
One-third of hospitalized patients are aged 65 years and older. In 1996, although
comprising only 13% of the US population they accounted for 38% of the approximately 31
million discharges from non-government, acute care hospitals. 7 Since all hospitalized older
patients are potentially at risk, the target population is quite large. Appropriate selection of
patients who are at risk for hospital-related complications and who are likely to receive benefit
from this practice, however, would decrease this number. 8
The target safety problems are preventable deaths and hospital-related functional decline
in older persons. The number of deaths that could be prevented if the practice were to be widely
implemented is unknown. On the other hand, since hospital-related functional decline occurs in
25 to 60% of older hospitalized patients, there is substantial opportunity to improve clinical
outcomes. 9 Other clinical problems explicitly studied in controlled trials include cognitive status
and nursing home placement.
Opportunities for Impact
Data from the American Hospital Association (AHA) 10 indicate that fewer than half of
hospitals providing care for the elderly have geriatric acute care units or offer comprehensive
geriatric assessment. (see footnote in Chapter 29). Researchers in the field believe the number is
increasing. The Department of Veterans Affairs reports that in 1997 there were 110 active GEM
units, although some concentrate solely on outpatient assessment. 11 A recent national survey 12
identified at least 15 active ACE units, with average daily censuses ranging from 5 to 25
patients. Depending on the screening and targeting criteria used to identify eligible patients, the
potential for impact could be quite large and raises the question of the physical and manpower
capacity of these units to meet the apparent demand.
Study Designs
A structured literature search identified a systematic review 13 that included 6 studies (4
randomized controlled trials, 14-17 one retrospective cohort study with historical controls, 18 and
one published abstract of a randomized controlled trial 19 ). We also identified 2 randomized
controlled trials of ACE units 5,20 that were published later (see Table 30.1). All of the cited
studies were single center in design. Five of the studies 5,14,16,17,20 provided sufficient data to
evaluate the baseline level of function of enrolled patients.
Study Outcomes
All-cause mortality was reported in each study. For this outcome, most patients were
followed 6 or more months after hospitalization. Other clinically important outcomes measured
in some studies were functional status, 5,14,16,17,20 cognitive status, 14,16,17 length of stay, 5,14-16,18,20
and discharge rates to institutional settings. 5,14-20
Evidence for Effectiveness of the Practice
In some studies mortality during hospitalization, at 3 months, or at 6 months was reduced
in the intervention group but the differences failed to achieve statistical significance. A metaanalysis
13 of 6 studies 14-19 found the summary odds ratio for 6-month mortality was 0.65 (95%
CI: 0.46-0.91), using both published and unpublished data from the included trials. Tests for
heterogeneity across studies were reported with p

Cognitive function, as measured by the Kahn-Goldfarb Mental Status Questionnaire, 21
showed no statistical improvement over the course of one study, 17 nor did 2 other studies 14,16
demonstrate improvement, using the Mini-Mental State Examination 22 to assess mental status.
Two trials of ACE units examined functional status, using the basic activities of daily living
(ADL). 23 Landefeld et al 5 reported statistically significant improvements, while Counsell et al 20
found benefit in a composite outcome of ADL improvement and nursing home placement, but
not in discharge ADL levels alone. Two other studies also demonstrated statistically improved
functional status in the six months after randomization 14 and at 12 months follow-up. 17
In individual studies, GEM units were associated with a higher likelihood of home
residence, rather than in an institutional setting (skilled nursing facility or nursing home).
325
5,14, 17
The meta-analysis by Stuck et al calculated a combined odds ratio that revealed a statistically
significant increase in patients discharged from GEM units who were living at home at 6 months
(summary odds ratio 1.80, 95% CI: 1.28-2.53) and 12 months (summary odds ratio 1.68, 95%
CI: 1.17-2.41) thereafter. 13 ACE unit trials in a community hospital 20 and a university hospital 5
were also both successful in decreasing patient placement in institutional settings and would
likely have strengthened the summary estimate if included. Extrapolating their study findings to
the US population, Rubenstein and colleagues 17 estimated that approximately 200,000 nursing
home admissions per year could be avoided using their geriatric evaluation unit approach.
Winograd noted that this multidisciplinary practice would potentially be more effective if target
populations were better identified and enrolled using specific criteria. 8
Potential for Harm
No data suggest that GEM units were associated with harm.
Costs and Implementation
Implementation of the practice requires construction or redesign of hospital ward(s) to
create a suitable environment, training or recruiting experienced staff, establishing selection
criteria to determine patient eligibility, and implementing a continuous evaluation process to
assess clinical and non-clinical outcomes.
A working group has recommended including costs as an important outcome measure in
future studies of GEM units. 24 Applegate and colleagues reported in a later analysis 25 of their
randomized controlled trial 14 that the increased costs associated with their intervention study
were not balanced by savings in subsequent health care spending, but if charges were adjusted
for days spent at home (versus in long-term care) the charges were similar. Lower direct costs
were demonstrated by one intervention study, 17 particularly after adjusting for differences in
survival (mean institutional-care costs per year survived, $22,597 for intervention patients vs.
$27,826 for control-group patients). The study by Landefeld et al 5 reported mean hospital
charges of $10,289 for intervention patients compared with $12,412 for control patients, with
similar median charges (p=0.3). Additional costs of about $75,000 attributable to staffing and
physical redesign of the unit resulted in a cost of about $230 per patient in the intervention
group. Counsell et al, 20 in a recent large randomized trial of an ACE intervention, reported no
difference in hospital costs for patients in the intervention group compared with the usual care
group ($5640 vs. $5754, respectively; p=0.68). Included in the costs was $28 per hospital day
per intervention patient, representing costs of the geriatric nurse specialist, unit medical director,
and unit renovations ($300,000).

Comment
Reasonable evidence supports the use of GEM units, despite varying findings across
individual studies with respect to their effectiveness at preventing outcomes such as mortality.
Nonetheless, mortality appears to be improved after pooling results of smaller trials. There is
good evidence that this model of care decreases nursing home placements, which in itself is a
noteworthy finding. Furthermore, the intervention costs may not be significantly higher in these
specialized units. The generalizability of the practice requires further examination, and the need
for multicenter studies, as advocated by a previous consensus group, 26 has thus far not been
undertaken.
Limitations of this practice compared with multidisciplinary geriatric consultation teams
(Chapter 29) include limited bed availability in most units, decreased transferability of geriatric
practices throughout a hospital, and a larger resource commitment compared with a hospitalwide
consultation team. The advantages of the GEM and ACE unit model are direct control over
implementation of clinical recommendations, the presence of dedicated geriatric nursing and
rehabilitative staff associated with the unit, and the beneficial effects of a ward designed to
address older patients’ needs. In sum, the practice of a dedicated GEM or ACE unit carries much
promise.
326

Table 30.1. Randomized Controlled Trials of Geriatric Evaluation and Management Units*
Study Study Setting Study
Design
Stuck,
1993 13
Applegat
e, 1990 14
Counsell,
2000 20
Gilchrist,
1998 15
Harris,
1991 16
Landefel
d, 1995 5
Powell,
1990 19
6 studies (published 1983-
1991) in the US, UK,
Australia, and Canada,
involving 1090 patients
(meta-analysis of Refs. 14-
19)
155 patients in a
university-affiliated
community hospital, 1985-
1987
1531 patients in a
community hospital, 1994-
1997
222 women on an
orthopedic-geriatric service
in the UK, 1984-1986
267 patients in an
Australian hospital, 1985-
1986
651 patients in a
university-affiliated
hospital, 1990-1992
203 patients in two
Canadian teaching
hospitals, year not stated
327
All-Cause Mortality and Other Outcomes†
Level 1A 6-month mortality: summary odds ratio 0.65
(95% CI: 0.46-0.91);
12-month mortality: summary odds ratio
0.77 (95% CI: 0.56-1.06)
Level 1 6-month mortality: 10% vs. 21% (p=NS);
After 6 months: greatest difference p=0.08
by log-rank test;
Improvement in ADLs: 3 of 8 ADLs better
in intervention group (p0.1)
Level 1 Inpatient mortality: estimated from Figure 1
in paper, 8% vs. 6% (p=NS);
12-month mortality: 23% vs. 29% (p=NS)
Level 1 Inpatient mortality: 7% in both groups
(p=NS);
3-month mortality: 14% vs. 13% (p=NS);
Improvement in ADLs at discharge: 34%
vs. 24% (p=0.009);
Discharged to nursing home: 14% vs. 22%
(p=0.01)
Level 1 Mortality: lower in intervention group;
timing not stated (p not stated);
Transferred to long-term care: fewer in
intervention group (p not stated)

Rubenstei
n, 1984 17
Teasdale,
1983 18
123 patients in a VA
hospital, 1980-1982
124 patients in a VA
hospital, 1981-1982
Level 1 Inpatient mortality: 14.3% vs. 15.0%
(p=NS);
12-month mortality: 23.8% vs. 48.3%
(p

References
1. Creditor MC. Hazards of hospitalization of the elderly. Ann Intern Med. 1993;118:219-223.
2. Hirsch CH, Sommers L, Olsen A, Mullen L, Winograd CH. The natural history of
functional morbidity in hospitalized older patients. J Am Geriatr Soc. 1990;38:1296-1303.
3. Stroke Unit Trialists Collaboration. How do stroke units improve patient outcomes? A
collaborative systematic review of the randomized trials. Stroke. 1997;28:2139-2144.
4. Palmer RM, Counsell S, Landefeld CS. Clinical intervention trials: the ACE unit. Clin
Geriatr Med. 1998;14:831-849.
5. Landefeld CS, Palmer RM, Kresevic DM, Fortinsky RH, Kowal J. A randomized trial of
care in a hospital medical unit especially designed to improve the functional outcomes of
acutely ill older patients. N Engl J Med.1995;332:1338-1344.
6. Palmer RM, Landefeld CS, Kresevic D, Kowal J. A medical unit for the acute care of the
elderly. J Am Geriatr Soc. 1994;42:545-552.
7. Kramarow E, Lentzner E, Rooks R, Weeks J, Saydah S. Health and aging chartbook.
Health, United States, 1999. Hyattsville, Maryland: National Center for Health Statistics,
1999.
8. Winograd CH. Targeting strategies: an overview of criteria and outcomes. J Am Geriatr
Soc. 1991;39S:25S-35S.
9. Palmer RM. Acute hospital care of the elderly: minimizing the risk of functional decline.
Cleve Clin J Med.1995;62:117-128.
10. American Hospital Association. Health Statistics, 2001 edition. Chicago, Illinois: Health
Forum, 2001.
11. Veterans Administration. VA programs for senior veterans. Geriatric evaluation and
management program. Available at: http://www.va.gov/seniors/health/gem.htm. Accessed
May 14, 2001.
12. Raziano DB, Jayadevappa R, Valenzula D, Weiner M, Lavizzo-Mourey RJ. Do Acute Care
for the Elderly (ACE) units work? Paper presented at: American Geriatrics Society Annual
Meeting; May 10, 2001; Chicago, IL. 2001.
13. Stuck AE, Siu AL, Wieland GD, Adams J, Rubenstein LZ. Comprehensive geriatric
assessment: a meta-analysis of controlled trials. Lancet. 1993;342:1032-1036.
14. Applegate WB, Miller ST, Graney MJ, Elam JT, Burns R, Akins DE. A randomized,
controlled trial of a geriatric assessment unit in a community rehabilitation hospital. N Engl
J Med. 1990;322:1572-1578.
15. Gilchrist WJ, Newman RJ, Hamblen DL, Williams BO. Prospective randomised study of
an orthopaedic geriatric inpatient service. BMJ. 1988;297:1116-1118.
16. Harris RD, Henschke PJ, Popplewell PY, et al. A randomised study of outcomes in a
defined group of acutely ill elderly patients managed in a geriatric assessment unit or a
general medical unit. Aust N Z J Med. 1991;21:230-234.
17. Rubenstein LZ, Josephson KR, Wieland GD, English PA, Sayre JA, Kane RL.
Effectiveness of a geriatric evaluation unit. A randomized clinical trial. N Engl J Med.
1984;311:1664-1670.
18. Teasdale TA, Shuman L, Snow E, Luchi RJ. A comparison of placement outcomes of
geriatric cohorts receiving care in a geriatric assessment unit and on general medicine
floors. J Am Geriatr Soc. 1983;31:529-534.
19. Powell C, Montgomery P. The age study: the admission of geriatric patients through
emergency [abstract]. Age & Ageing. 1990;19(Suppl. 2):P21-P22.
329

20. Counsell SR, Holder CM, Liebenauer MA, et al. Effects of a multicomponent intervention
on functional outcomes and process of care in hospitalized older patients: a randomized
controlled trial of Acute Care for Elders (ACE) in a community hospital. J Am Geriatr Soc.
2000;48:1572-1581.
21. Kahn RL, Goldfarb AI, Pollack M, Peck A. Brief objective measures for the determination
of mental states in the aged. Am J Psychiatry. 1960;117:326-328.
22. Folstein MF, Folstein SE, McHugh PR. "Mini-mental state": a practical method for grading
the cognitive state of patients by the clinician. J Psychiatr Res. 1975;12:189-198.
23. Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of illness in the aged.
The index of ADL: a standardized measure of biological and psychosocial function. JAMA.
1963;185:914-919.
24. Hedrick SC, Barrand N, Deyo R, et al. Working group recommendations: measuring
outcomes of care in geriatric evaluation and management units. J Am Geriatr Soc.
1991;39S:48S-52S.
25. Applegate WB, Graney MJ, Miller ST, Elam JT. Impact of a geriatric assessment unit on
subsequent health care charges. Am J Public Health. 1991;81:1302-1306.
26. Kramer A, Deyo R, Applegate W, Meehan S. Research strategies for geriatric evaluation
and management: conference summary and recommendations. J Am Geriatr Soc.
1991;39S:53S-57S.
330

Section E. General Clinical Topics
Chapter 31. Prevention of Venous Thromboembolism
Chapter 32. Prevention of Contrast-Induced Nephropathy
Chapter 33. Nutritional Support
Chapter 34. Prevention of Clinically Significant Gastrointestinal Bleeding in Intensive Care
Unit Patients
Chapter 35. Reducing Errors in the Interpretation of Plain Radiographs and Computed
Tomography Scans
Chapter 36. Pneumococcal Vaccination Prior to Hospital Discharge
Chapter 37. Pain Management
Subchapter 37.1. Use of Analgesics in the Acute Abdomen
Subchapter 37.2. Acute Pain Services
Subchapter 37.3. Prophylactic Antiemetics During Patient-controlled Analgesia
Therapy
Subchapter 37.4. Non-pharmacologic Interventions for Postoperative Plan
331

332

Chapter 31. Prevention of Venous Thromboembolism
Jennifer Kleinbart, MD
Mark V. Williams, MD
Kimberly Rask, MD, PhD
Emory University Schools of Medicine and Public Health
Background
Venous thromboembolism (VTE) refers to occlusion within the venous system. It
includes deep vein thrombosis (DVT), typically of the lower extremities, and embolism to the
pulmonary vasculature. Distal DVT is occlusion confined to the deep calf veins, while
thrombosis at or above the popliteal vein is considered proximal DVT. A distal DVT becomes
clinically important if it extends proximally, where the chance of pulmonary embolization (PE)
is clinically significant.
In hospitalized patients, VTE occurs with relatively high frequency. A patient’s risk of
VTE varies depending on multiple factors including age, medical condition, type of surgery,
duration of immobilization, and the presence of an underlying hypercoagulable state such as
malignancy. Measures to prevent VTE have been widely studied for many reasons, including
VTE’s high incidence, its associated mortality and morbidity, its cost of treatment, and its
treatment-related complications.
VTE is often clinically silent. As a result, studies evaluating the efficacy of preventive
measures generally screen patients who are asymptomatic. As widespread screening is not
recommended in general practice, the incidence of VTE in most studies appears higher than that
encountered in clinical practice. The importance of clinically undetected VTE is not fully
understood.
Practice Description
Both mechanical and pharmacologic interventions have been evaluated for prevention of
VTE (Table 31.1). Mechanical devices include graduated elastic stockings (ES) and intermittent
pneumatic compression (IPC); pharmacologic measures include low dose unfractionated heparin
(LDUH), low molecular weight heparin (LMWH), warfarin, and aspirin .
Prevalence and Severity of the Target Safety Problems
There are over 23 million surgeries performed each year in the United States. 1 The
frequency of DVT and PE varies by type of procedure and specific patient risk factors, as set
forth below. In general, without prophylaxis DVT occurs after approximately 20% of all major
surgical procedures and PE in 1-2%. Over 50% of major orthopedic procedures are complicated
by DVT and up to 30% by PE if prophylactic treatment is not instituted. 2
In addition, of the more than 31 million patients admitted each year for medical
conditions, up to 16% of patients will develop DVT in the absence of prophylaxis. 1,2
333

Opportunities for Impact
Despite the frequency with which VTE occurs in hospitalized patients, and the wellestablished
efficacy and safety of preventative measures, prophylaxis is often underused or used
inappropriately. One survey of general surgeons found that 14% did not use VTE prophylaxis. 3
Another survey of orthopedic surgeons found that only 55% placed all hip fracture patients on
VTE prophylaxis, and 12% never used prophylaxis. 4 When performing total hip replacement
(THR) and total knee replacement (TKR), 81-84% of surgeons placed all patients on
prophylaxis, and 3-5% used no prophylaxis. Taken together, these data imply that VTE
prophylaxis is used in 92% of THR patients, 89% of TKR patients, and 73% of hip fracture
repairs. For spinal procedures, 21-38% of surgeons used prophylaxis with all patients, while 46-
64% did not use prophylaxis with any patients, an overall rate of prophylaxis of 25-44%. 4 A
more recent chart review of Medicare patients over age 65 undergoing major abdominothoracic
surgery from 20 Oklahoma hospitals found that only 38% of patients were given VTE
prophylaxis. 5 Of patients considered at very high risk for VTE, the same percentage received
some form of prophylaxis, but only 66% of those received appropriate preventive measures. 5
Another chart review of high-risk patients at 10 acute care US hospitals found that some
type of VTE prophylaxis was used in 94% of orthopedic patients and 75% of abdominal surgery
patients. 6 However, the 1995 American College of Chest Physician grade A recommendations 7
(those therapies supported by the most rigorously designed studies) were followed in only 84%
of THRs, 76% of TKRs, 45% of hip fracture repairs, and 50% of abdominal surgery patients.
Compliance with grade A recommendations was no better in the 3 hospitals that had critical
pathways, pocket guides, or a policy for VTE prophylaxis. 6
Inadequate VTE prophylaxis among medical patients may be even more prevalent. In one
study, only 33% of patients admitted to a medical intensive care unit received VTE prophylaxis.
Eighty-seven percent of these patients had one risk factor, and over 50% had multiple risks for
VTE. 8 Another study found that of patients who developed VTE during admission or within 30
days of discharge, 48% had not received prophylaxis during hospitalization. Most of these were
medical patients. 9
Study Designs and Outcomes
There are a large number of randomized control trials (RCTs) and high quality metaanalyses
that examine the efficacy of VTE prophylaxis. Most studies considered surgical rather
than medical patients. Several differences in study design may account for heterogeneity among
these studies, including differences in patients or procedures, the intervention (type, duration, or
dose of prophylaxis), the method used to diagnosis VTE, the outcome measured (distal or
proximal DVT, fatal or nonfatal PE), and whether the endpoint was a clinical event or one found
by routine screening.
The “gold standard” for diagnosis of DVT is contrast venography, which is generally
used in studies screening high-risk patients. Venography may detect a significant number of clots
that are not clinically important, and is often technically limited. 10 Other common screening
methods are fibrinogen leg scanning (fibrinogen uptake test, FUT) and duplex ultrasonography,
both of which have low sensitivity for calf vein thrombosis. 11,12 Meta-analysis of studies using
the FUT in orthopedic patients showed a sensitivity of 55% for calf vein thrombosis and 45% for
all DVT, with a specificity of 92%. 11 For detection of calf vein thrombosis, duplex
ultrasonography had a sensitivity and specificity of 39% and 98% in symptomatic hospitalized
patients, and of 13% and 92% in patients undergoing arthroplasty. 12 In studies of general surgery
334

patients, the incidence of DVT was 25% when diagnosis was made by FUT, and 19% when
confirmed by venogram. 2 However, in trials of VTE prophylaxis using elastic stockings in
neurosurgical patients, the study using FUT found a DVT rate of 9%, while 3 others using
venography found the rate of DVT to be 28%. 2
In many studies, all patients are screened for DVT, although several studies only tested
for DVT if clinical signs or symptoms occurred. Clearly, in the latter a number of DVT are
missed; however, the clinical importance of these asymptomatic clots is unknown.
The gold standard for diagnosis of pulmonary embolus is pulmonary angiography.
However, reliable evidence confirms that combining clinical probability with results of
ventilation-perfusion scanning is an accurate and less invasive method of making this
diagnosis. 13 Most studies report the incidence of symptomatic or fatal PE. Studies rarely screen
for PE in asymptomatic patients. Silent pulmonary embolism may occur in up to 50% of patients
with proximal DVT, however, the clinical importance of these emboli is unclear. 14
Evidence for Effectiveness of the Practice
Studies of VTE prophylaxis are best grouped by the population at risk (Table 31.2). The
sections that follow describe the prevalence of thromboembolism in patients in each category
and discuss the efficacy of various prophylactic strategies.
General Surgery
For general surgery patients not receiving VTE prophylaxis, the incidence of DVT
confirmed by venogram is 19%. Proximal DVT occurs in 7% of patients, PE in 1.6%, and fatal
PE in 0.9%. 2
In general surgery patients, the risk of VTE is highest in those undergoing major surgery
who are over the age of 40 and have an underlying hypercoagulable state, prior VTE, or cancer.
(Patients with spinal cord injury, trauma, and those undergoing hip or knee arthroplasty or hip
fracture surgery also fall in this risk group and are discussed separately below). Incidence rates
for patients in this “very high” risk group are as follows: calf DVT, up to 80%; proximal DVT,
20%; PE, 10%; and up to 5% may suffer a fatal PE if no prophylaxis is provided. 2 The risk of
VTE is lowest in those surgical patients who are young, otherwise healthy, and whose surgeries
are “minor.” Such patients have incident rates as follows: calf DVT, 2%; proximal DVT, 0.4%,
PE , 0.2%; and fatal PE , 0.002%.
Numerous studies show that both LDUH and LMWH reduce the risk of proximal DVT,
PE, and fatal PE in patients undergoing general surgery. Pooled results from 46 randomized
trials show that prophylaxis of general surgical patients with LDUH compared with placebo
reduced the risk of DVT (diagnosed by FUT or FUT confirmed with venography) by 68%, (from
25% [95% CI: 24-27%] to 8% [95% CI: 7-8%]). 2 LMWH has comparable efficacy to LDUH for
prevention of VTE, and may be more effective for preventing proximal DVT and PE. 2,15,16
Pooled results (though heterogeneity was present) from 26 studies showed that high dose
LMWH (>3400 anti-Xa units) does not reduce DVT more than low dose LMWH does (

Orthopedic Patients
All studies included in the evaluation of prophylaxis for orthopedic patients used
venography to diagnose DVT.
Total Hip Replacement
Within 7-14 days after total hip replacement (THR), patients not receiving prophylaxis
have an incidence of total and proximal DVT of 54% and 25%, respectively. Asymptomatic PE
may occur in 5-15%, symptomatic PE in 1-2%, and fatal PE in 0.1-0.4%. 2,21 In patients
undergoing THR, many RCTs have shown that both mechanical and pharmacologic measures
are highly effective in the prevention of VTE.
A meta-analysis of 52 RCTs showed that patients receiving prophylaxis with LMWH,
LDUH, warfarin, aspirin or IPC had fewer total DVTs (proximal plus distal). 21 IPC and ES
reduce distal DVT but do not significantly reduce proximal DVT in these patients. 2,17 Compared
with placebo, prophylaxis with warfarin or LMWH resulted in the greatest reduction in proximal
DVT: a reduction of 70-80% with either method (risk of proximal DVT with warfarin 5-6%, RR
6.3% [95% CI: 4.7-8.4%]; with LMWH 6–8%, RR 7.7% [95% CI: 5.7-10.3%]). 2,21 Both LMWH
and warfarin resulted in significantly fewer proximal DVTs compared with LDUH or IPC
(p

data show that aspirin (6 studies, 443 patients), warfarin (9 studies, 1294 patients), and LDUH (2
studies, 172 patients) do not significantly reduce proximal DVT following TKR. 2 Patients given
warfarin prophylaxis who were routinely screened with venography had an 8–12% risk of
proximal DVT. However, one study that followed patients for 3 months following TKR found
that patients who received warfarin had a rate of symptomatic VTE of only 0.8%. 24
The incidence of PE is low in TKR patients. Studies in which PE was diagnosed by lung
scan or angiography showed the rate of symptomatic and asymptomatic PE in patients treated
with aspirin to be 1.3% and 11.7%, respectively, with warfarin 0.4% and 8.2%, and with IPC
0.5% and 6.3%. Symptomatic PE occurred in none of 177 patients receiving LMWH. No studies
of LMWH used routine lung scanning. 23
Initiation of Therapy
Meta-analysis of patients undergoing THR found that LMWH initiated preoperatively
resulted in a lower risk of VTE than LMWH started postoperatively (10% vs. 15%, p=0.02). 25
Major bleeding occurred less frequently in the group receiving preoperative LMWH (0.9% vs.
3.5%, p=0.01).
Duration of Therapy
The appropriate duration of VTE prophylaxis following orthopedic surgery is not clearly
established. However, it is clear that the increased risk of DVT persists post-discharge. In the
largest randomized trial of post-discharge LMWH, 533 patients received either 35 days of
dalteparin or 6 days of warfarin followed by placebo. Patients treated with extended LMWH had
significantly fewer total and proximal DVTs from day 6 to day 35 (for all DVT, dalteparin 4.8%
vs. placebo 10.5%, p=0.03; for proximal DVT, dalteparin 1% vs. placebo 4.8%, p=0.02), as well
as from day 1 to day 35 (dalteparin 2.6% vs. warfarin/placebo 9.2%, p=0.002). Seventy five
percent of these DVTs were asymptomatic. Symptomatic thrombi occurred in approximately 1-
2%. No patient had symptomatic, objectively documented pulmonary embolism. 26 Pooled data
from this study and 5 others that compared in-hospital LMWH followed by LMWH or placebo
found that prolonged LMWH resulted in a 66% reduction in total DVT (14% vs. 27%) and a
66% reduction in proximal DVT (4% vs. 12%) by 35 days. 2
Neurosurgery
Patients undergoing neurosurgical procedures carry a risk of developing DVT of
approximately 22%, and a risk of proximal DVT of 5%. 2 The risk of DVT is increased in
patients undergoing intracranial surgery compared to spinal surgery. Among patients undergoing
intracranial surgery, those with malignant tumors have a higher risk of DVT than those with
benign tumors. Increasing age and increasing duration of neurosurgery further increase the risk
of VTE. Meta-analysis of randomized studies comparing LMWH to placebo (one study) or
LMWH plus ES to ES alone (2 studies) in neurosurgical patients found that LMWH was
associated with a 38% reduction in total DVT (18% vs. 28%, p

Trauma
Trauma patients, especially those with orthopedic injuries, are at very high risk for VTE.
DVT occurs in over 50% of these patients, proximal DVT in approximately 20%, and fatal PE in
up to 2%. 2 Few randomized trials have studied VTE prophylaxis in trauma patients. Metaanalysis
shows that VTE prophylaxis, either with LDUH (4 randomized trials, OR 0.97, 95% CI:
0.35-2.6 for LDUH vs. control) or ES (3 randomized trials, OR 0.77, 95% CI: 0.27-2.2), did not
reduce DVT in trauma patients compared with placebo. 29 Heterogeneity was present for all
comparisons due to differences in methods used to diagnose DVT and differences in trauma
populations. Pooled data from 4 studies (2 randomized and 2 nonrandomized) of LDUH versus
mechanical prophylaxis found no difference in risk of DVT (OR 1.16, 95% CI: 0.5-2.7). 29 One
randomized trial comparing LDUH with LMWH in trauma patients screened with venography
found that DVT was reduced by 30% with enoxaparin (31% vs. 44%, p=0.01). 30 Of the 265
patients randomized, only one patient (in the LMWH group) suffered a PE.
Acute Spinal Cord Injury
The risk of DVT (diagnosed by FUT or impedance plethysmography) in patients with
acute spinal cord injury ranges from 40-90%. 2 The only trial using screening venography in
patients with acute spinal cord injury who were not receiving prophylaxis found DVT in 81%
and proximal DVT in 35%. 31
Prophylaxis in acute spinal cord injury has not been well studied. In the only study
evaluating use of IPC, proximal DVT occurred in 40% of patients with IPC alone, and in 25% of
patients with IPC combined with aspirin and dipyridamole. 32 One randomized trial with 35
patients found that LMWH (tinzaparin 3500U daily) compared to LDUH (5000U three times a
day) reduced DVT from 16% to 0%, a reduction that did not reach statistical significance. 33
However, a later prospective cohort study diagnosed DVT in 13% of 48 patients prophylactically
treated with LMWH. 34 In both studies, screening for DVT was done with ultrasound or
ultrasound plus impedance plethysmography, with confirmation by venography. Of 15 acute
spinal cord injury patients from a randomized trial of major trauma patients, DVT was detected
in 67% of those given LDUH and 50% with LMWH. Proximal DVT occurred in 13% receiving
LDUH and none prophylaxed with LMWH. 30
338

Medical Patients
VTE prevention in hospitalized medical patients has not been studied as extensively as in
surgical patients. DVT occurs in 24% of patients with myocardial infarction, and is reduced by
71% with LDUH compared with placebo with no increase in bleeding (4 studies with 165
patients). 2 However, most patients with acute myocardial infarction receive full anticoagulation
for treatment of the acute coronary syndrome.
After ischemic stroke, 55% of untreated patients develop DVT. Prophylaxis with either
LDUH or LMWH given for 10-14 days reduces DVT by 56%, from 55% (95% CI: 49%-60%) to
24% (95% CI: 20%-29%). Two studies directly comparing LDUH (5000 U three times daily) to
LMWH (enoxaparin 40 mg once daily), using venography for diagnosis, found greater reduction
in DVT with LMWH. 2
Among hospitalized patients with other medical conditions, the rate of DVT is
approximately 16%. 2 A meta-analysis of studies of hospitalized patients with conditions other
than myocardial infarction or ischemic stroke given VTE prophylaxis with unfractionated or low
molecular weight heparin showed a 56% reduction in DVT (RR 0.44, 95% CI: 0.29-0.64) and a
52% reduction in PE (RR 0.48, 95% CI: 0.34-0.68). No significant difference was found between
LMWH and LDUH in incidence of DVT, PE, or mortality; however, major hemorrhage was
lower with LMWH than with LDUH (RR 0.48, 95% CI: 0.23-1.00). 35
Another randomized trial comparing enoxaparin with placebo in 866 medical patients
found a 63% reduction in overall VTE risk with enoxaparin (40 mg) compared with placebo in
the first 14 days (5.5% vs. 15%, RR 0.37, 95% CI 0.22-0.63). Proximal DVT was reduced by
65% (1.7% vs. 4.9%, RR 0.40, 95% CI 0.23-0.69). Significant reductions in total and proximal
DVT persisted at 110 days. VTE was not reduced in the group receiving 20 mg of enoxaparin.
The most common medical conditions were acute infectious disease, acute respiratory failure,
New York Heart Association Class III or IV congestive heart failure. 36
The risk of VTE is higher in patients with malignant disease, especially those with
adenocarcinoma or brain tumors. Factors associated with increased VTE in cancer patients
include chemotherapy, surgery, and indwelling central venous catheters. Breast cancer patients
treated with tamoxifen also have higher rates of VTE. In patients with indwelling central venous
catheters who received low dose warfarin (1 mg per day), upper extremity DVT was reduced by
75% (9.5% vs. 37.5%). 37
In summary, for general surgery patients who are at moderate risk or greater for VTE,
LDUH and LMWH are the most effective methods for prophylaxis. IPC also provides effective
DVT prevention, but in very high risk patients should only be used in conjunction with heparin.
For orthopedic procedures, LMWH and warfarin are the most effective preventive measures.
Neurosurgical patients should receive LMWH or LDUH, while in acute spinal cord injury or
trauma, LMWH provides the largest reduction in VTE. For stroke or medical patients, LMWH
and LDUH show the greatest benefit for VTE prevention (see Table 31.2).
Summary of Prophylaxis Recommendations
For general surgery patients at moderate to high risk, LDUH and LMWH have
comparable effectiveness for prevention of VTE, with similar bleeding risks. LDUH is generally
more cost-effective. For high risk patients, IPC has not been consistently shown to prevent
proximal DVT (See Table 31.3).
For major orthopedic procedures (THR, TKR, hip fracture repair), prophylaxis with
LMWH or warfarin results in the greatest benefit, with a small increase in bleeds compared with
339

no prophylaxis, but no difference between the two agents. Warfarin may be more cost-effective,
but necessitates monitoring and dose adjustment. This should be considered in choosing between
the two agents.
IPC, LMWH, and LDUH are all acceptable methods of prophylaxis for neurosurgical
patients. Each effectively reduces proximal DVT with no increase in major bleeding. There are
no data on cost-effectiveness of prophylaxis for these patients.
Data for prophylaxis of trauma patients does not show conclusive VTE reduction with
any agent. However, the risk of VTE among these patients is high and prophylaxis should be
considered, especially for those with orthopedic injuries. LMWH is a safe method of prophylaxis
and has not been shown to increase major bleeds in blunt trauma patients. However if a patient is
at high risk for major bleeding, IPC should be considered.
For medical patients, LDUH and LMWH are both effective for reducing VTE. LDUH
may result in slightly more bleeding, but is more cost-effective. Patients at high risk for bleeding
should be given prophylaxis with ES or IPC.
Potential for Harm
There is no documented risk associated with mechanical devices, although there is a
potential risk that patients’ legs will be examined less frequently. The major risk of
pharmacologic prophylaxis is bleeding. Bleeding is typically considered major when the
hemoglobin decreases by at least 2 g/dL, when red blood cell transfusion is necessary, when
intracranial or retroperitoneal bleeding occurs, or when bleeding requires surgical intervention.
Another consequence of heparin therapy may be thrombocytopenia.
For general surgery, there is no difference in the risk of major bleeding between LMWH
and LDUH, although fewer minor bleeds may occur with LMWH (RR 0.64, 95% CI 0.58-
0.70). 15 However, when evaluated according to dose of LMWH, compared with LDUH, lowdose
LMWH (

LMWH was more cost-effective than LDUH. 39 An economic evaluation using decision analysis
compared LDUH and LMWH for patients undergoing colorectal surgery. There was no
difference in risk of VTE between groups. Per 1000 patients treated, prophylaxis with
enoxaparin compared with LDUH resulted in 12 excess major bleeds and an additional cost of
$145,667. 40 This supports LDUH as a more cost-effective measure for patients undergoing
general surgery.
Meta-analysis of studies of the cost-effectiveness of VTE prophylaxis for patients
undergoing hip arthroplasty found that with a 2.6 to 1 price ratio between LMWH and LDUH,
use of LMWH would save the health care system approximately $50,000 per 1000 patients
treated. 41 For VTE prophylaxis after TKR, LMWH was more cost-effective than warfarin, saving
$2401 per 100 patients ($9197 vs. $11,598 per 100 patients). 42 This study was done in Canada,
where LMWH is less costly than in the United States. Another meta-analysis of THR patients
found that LDUH would decrease the cost of care related to DVT by $200 per patient. Compared
with warfarin, LMWH would be more effective in preventing DVT (expected DVT rate
420/10,000 with warfarin and 250/10,000 with LMWH), and death from VTE (110/10,000 with
warfarin and 70/10,000 with LMWH). However, preventing one death with LMWH use instead
of warfarin would cost approximately $12,000. 43
In medical patients, LDUH given 3 times daily was found to be more cost-effective than
LMWH, with a savings per 1000 patients of $10,753 compared with enoxaparin 40 mg/d, and
$15,000 compared with enoxaparin 20 mg/d. 44 The higher cost associated with enoxaparin 20
mg/d results from the higher incidence of complications with this regimen. This supports use of
LDUH as the preferred method of prophylaxis for medical patients at the present time, though
this is an area of active investigation. No studies were found evaluating the cost of DVT
prophylaxis in neurosurgery patients.
Comment
As noted earlier, despite the clear evidence of effectiveness, DVT prophylaxis is
underused. The reasons for this underuse have not been completely elucidated, nor have the
optimal strategies for improving prophylaxis been fully identified. Various strategies have been
tried in an effort to improve physician utilization of appropriate VTE prophylaxis. One hospital
studied the impact of educational programs promoting guidelines for prophylaxis. It found that
presentations to staff, distributions of cards with the hospital’s guidelines, and posters increased
prophylaxis of surgical patients from 59% to 70%, and to 77% for high-risk patients (see also
Chapters 51 and 54). 45
In another study, a computer-based clinical decision support system (CDSS) (see Chapter
53) providing information pertaining to VTE prophylaxis was used in an orthopedic surgery
department of a teaching hospital. The investigators monitored the impact of CDSS use in 1971
patients undergoing orthopedic surgery. Compliance with guidelines was 83% during control
periods and 95% during intervention periods. Inappropriate practice decisions occurred almost 4
times more frequently during control versus intervention periods. 46
A third study evaluating methods to improve VTE prophylaxis among intensive care unit
patients found that staff education improved use of appropriate prophylaxis from 38% to 62%.
When required computerized order sets were added to education, appropriate prophylaxis
increased to 97%. 47
These studies suggest that either a knowledge gap or lack of awareness may exist among
practitioners. For institutions or groups attempting to improve appropriate use of measures to
prevent VTE, guidelines made available via computerized support systems or order sets provide
341

the most effective means of implementing appropriate VTE prophylaxis, especially when these
systems are linked to effective educational programs.
342

Table 31.1. Mechanical and pharmacologic preventative measures for VTE
Practice Type Description Comment
Graduated Elastic
Stockings (ES)
Intermittent
pneumatic
compression (IPC)
Mechanical Fitted hose that extend
above the knee
Mechanical Devices fitted over lower
extremities that
sequentially inflate and
deflate
Aspirin Pharmacologic Usually 325 mg/d
Warfarin Pharmacologic 5-10 mg started the day
of or after surgery; adjust
to achieve an INR of 2-3
Low-dose
unfractionated
heparin (LDUH)
Low Molecular
Weight Heparin
(LMWH)
Pharmacologic Generally 5000 U
subcutaneous bid or tid,
though some studies have
adjusted dose to maintain
PTT at high end of
normal
Pharmacologic Dose depends on type of
surgery and VTE risk*
343
Fitted hose are more
efficacious than non-fitted
Monitoring of INR needed
Contraindicated if active
bleeding or history of
thrombocytopenia; no
need to follow coagulation
studies (unless adjusted
dose is used)
No need to monitor
coagulation studies
* LMWH dosing: Enoxaparin 20 mg SC daily (moderate risk surgery) or 40 mg SC daily (can go
up to 30 mg SC q12h for high risk general surgery, major trauma or acute spinal cord injury);
dalteparin 2500–5000 U SC daily; nadroparin 2500 U SC daily; tinzaparin 3500-4500 U SC
daily (may be dosed 75U/kg/d for orthopedic surgery).

Table 31.2. Summary of DVT risk and prophylactic methods providing significant risk
reduction*
Surgery/
Condition
General
Surgery 2
THR 2
TKR 2,23
Neurosurgery
27,28
Trauma 2,30
Risk of all DVT in
untreated patients
25%
54%
64%
28%
Type of
Prophylaxis
30-60% LMWH
Acute Spinal
Cord Injury 2 80% Not established
Ischemic
stroke 2
55%
344
Risk Reduction with
Prophylaxis
ES 44% 3
LDUH 68% 47
LMWH 76% 21
IPC 88% 2
LMWH 70% 30
warfarin 59% 13
LMWH 52% 13
IPC 73% 6
LMWH 38% 3
LDUH 72% 1†
30%
(compared to LDUH)
LDUH 56% 5
LMWH 58% 3
Danaparoid 82% 4
Number of
Studies
76% 2†
Medical
LMWH
39% 2
conditions 2 16%
LDUH 61% 3†
* DVT indicates deep venous thrombosis; ES, graduated elastic stockings; IPC, intermittent
pneumatic compression; LDUH, low-dose unfractionated heparin; LMWH, low molecular
weight heparin; THR, total hip replacement; and TKR, total knee replacement.
† DVT diagnosed by fibrinogen uptake test (FUT)
1

Table 31.3. Recommended VTE prophylaxis for surgical procedures and medical
conditions*
Surgery/Condition Recommended
Prophylaxis
General Surgery – low-risk: minor
procedures,

References
1. US Bureau of the Census. Statistical Abstract of the United States: 2000. 120th edition.
Washington, DC. 2000.
2. Geerts WH, Heit JA, Clagett GP, Pineo GF, Colwell CW, Anderson FA, et al. Prevention
of venous thromboembolism. Chest. 2001;119:132S-175S.
3. Caprini JA, Arcelus JI, Hoffman K, Mattern T, Laubach M, Size GP, et al. Prevention of
venous thromboembolism in North America: results of a survey among general surgeons. J
Vasc Surg. 1994;20:751-758.
4. Janku GV, Paiement GD, Green HD. Prevention of venous thromboembolism in
orthopaedics in the United States. Clin Ortho & Related Research. 1996:313-321.
5. Bratzler DW, Raskob GE, Murray CK, Bumpus LJ, Piatt DS. Underuse of venous
thromboembolism prophylaxis for general surgery patients: physician practices in the
community hospital setting. Arch Intern Med. 1998;158:1909-1912.
6. Stratton MA, Anderson FA, Bussey HI, Caprini J. Prevention of venous thromboembolism:
adherence to the 1995 American College of Chest Physicians Consensus Guidelines for
Surgical Patients. Arch Intern Med. 2000;160:334-340.
7. Clagett GP, Anderson FAJ, Levine MN, Wheeler HB. Prevention of venous
thromboembolism. Chest. 1995;108S:312S-334S.
8. Keane MG, P IE, Goldhaber SZ. Utilization of venous thromboembolism prophylaxis in
the medial intensive care unit. Chest. 1994;106:13-14.
9. Goldhaber SZ, Dunn K, MacDougall RC. New onset of venous thromboembolism among
hospitalized patients at Brigham and Women's Hospital is caused more often by
prophylaxis failure than by withholding treatment. Chest. 2000;118:1680-1684.
10. Heijboer H, Cogo A, Buller HR. Detection of deep vein thrombosis with impedance
plethysmography and real-time compression ultrasonography in hospitalized patients. Arch
Intern Med. 1992;152:1901-1903.
11. Lensing AW, Hirsh J. 1251-fibrinogen leg scanning: reassessment of its role for the
diagnosis of venous thrombosis in post-operative patients. Thromb Heamost. 1993;69:2-7.
12. Eskandari MK, Sugimoto H, Richardson T. Is color-flow duplex a good diagnostic test for
detection of isolated calf vein thrombosis in high-risk patients? Angiology. 2000;51:705-
710.
13. Perrier A, Miron MJ, Desmarais S, de Moerloose P. Using clinical evaluation and lung
scan to rule out suspected pulmonary embolism. Arch Intern Med. 2000;160:512-516.
14. Meignan M, Rosso J, Gauthier H, Brunengo F. Systematic lung scans reveal a high
frequency of silent pulmonary embolism in patients with proximal deep venous
thrombosis. Arch Intern Med. 2000;160:159-164.
15. Palmer AJ, Schramm W, Kirchhof B, Bergemann R. Low molecular weight heparin and
unfractionated heparin for prevention of thrombo-embolism in general surgery: a metaanalysis
of randomised clinical trials. Haemostasis. 1997;27:65-74.
16. Koch A, Bouges S, Ziegler S, Dinkel H, Daures JP, Victor N. Low molecular weight
heparin and unfractionated heparin in thrombosis prophylaxis after major surgical
intervention: update of previous meta-analyses. Br J Surg. 1997;84:750-759.
346

17. Vanek VW. Meta-analysis of effectiveness of intermittent pneumatic compression devices
with a comparison of thigh-high to knee-high sleeves. American Surgeon. 1998;64:1050-
1058.
18. Wells PS, Lensing AW, Hirsh J. Graduated compression stockings in the prevention of
postoperative venous thromboembolism. A meta-analysis. Arch Intern Med. 1994;154:67-
72.
19. Agu O, Hamilton G, Baker D. Graduated compression stockings in the prevention of
venous thromboembolism. Br J Surg. 1999;86:992-1004.
20. Amarigiri SV, Lees TA. Elastic compression stockings for prevention of deep vein
thrombosis. In: The Cochrane Library, Issue1, 2001.
21. Freedman KB, Brookenthal KR, Fitzgerald RH, Jr., Williams S, Lonner JH. A metaanalysis
of thromboembolic prophylaxis following elective total hip arthroplasty. J Bone
Joint Surg. 2000;82-A:929-938.
22. Pulmonary Embolism Prevention (PEP) Trial Collaborative Group. Prevention of
pulmonary embolism and deep vein thrombosis with low dose aspirin: Pulmonary
Embolism Prevention (PEP) trial. Lancet. 2000;355:1295-1302.
23. Westrich GH, Haas SB, Mosca P, Peterson M. Meta-analysis of thromboembolic
prophylaxis after total knee arthroplasty. J Bone Joint Surg. 2000;82-B:795-800.
24. Heit JA, Berkowitz SD, Bona R, Cabanas V, Corson JD, Elliott CG, et al. Efficacy and
safety of low molecular weight heparin compared to warfarin for the prevention of VTE
after total knee replacement surgery: a double-blind, dose-ranging study. Thromb Haemost.
1997;77:32-38.
25. Hull RD, Brant RF, Pineo GF, Stein PD. Preoperative vs postoperative initiatin of lowmolecular-weight
heparin prophylaxis against venous thromboembolism in patients
undergoing elective hip replacement. Arch Intern Med. 1999;159:137-141.
26. Hull RD, Pineo GF, Francis C, Bergqvist D, Fellenius C, Soderberg K, et al. Lowmolecular-weight
heparin prophylaxis using dalteparin extended out-of-hospital placebo in
hip arthroplasty patients. Arch Intern Med. 2000;160:2208-2215.
27. Iorio A, Agnelli G. Low-molecular-weight and unfractionated heparin for prevention of
venous thromboembolism in neurosurgery: a meta-analysis. Arch Intern Med.
2000;160:2327-2332.
28. Cerrato D, Ariano C, Fiacchino F. Deep vein thrombosis and low-dose heparin prophylaxis
in neurosurgical patients. J Neurosurg. 1978;49:378-381.
29. Velmahos GC, Kern J, Chan LS, Oder D, Murray JA, Shekelle P. Prevention of venous
thromboembolism after injury: an evidence-based report–part I: analysis of risk factors and
evaluation of the role of vena caval filters. J Trauma. 2000;49:132-138.
30. Geerts WH, Jay RM, Code KI. A comparison of low dose heparin with low molecular
weight heparin as prophylaxis against venous thromboembolism after major trauma. N
Engl J Med. 1996;335:701-707.
31. Geerts WH, Code KI, Jay RM, Chen E. A prospective study of venous thromboembolism
after major trauma. N Engl J Med. 1994;331:1601-1606.
32. Green D, Rossi EC, Yao JS. Deep vein thrombosis in spinal cord injury: effect of
prophylaxis with calf compression, aspirin and dipyridamole. Paraplegia. 1982;20:227-
234.
33. Green D, Lee MY, Lim AC. Prevention of thromboembolism after spinal cord injury using
low molecular weight heparin. Ann Intern Med. 1990;113:571-574.
347

34. Green D, Chen D, Chmiel JS. Prevention of thromboembolism in spinal cord injury: role of
low molecular weight heparin. Arch Phys Med Rehabil. 1994;75:290-292.
35. Mismetti P, Laporte-Simitsidis S, Tardy B, Cucherat M, Buchmuller A, Juillard-Delsart D,
et al. Prevention of venous thromboembolism in internal medicine with unfractionated or
low-molecular-weight heparins: a meta-analysis of randomised clinical trials. Thromb
Haemost. 2000;83:14-19.
36. Samama MM, Cohen AT, Darmon JY, Desjardins L, Eldor A, Janbon C, et al. A
comparison of enoxaparin with placebo for the prevention of VTE in acutely ill medical
patients. N Engl J Med. 1999;341:793-800.
37. Bern MM, Lokich JJ, Wallach SR. Very low doses of warfarin can prevent thrombosis in
central venous catheters: a randomized prospective trial. Ann Intern Med. 1990;112:423-
428.
38. Norwwod SH, McAuley CE, Berne JD, Vallina VL, Kerns DB, Grahm TW, et al. A
potentially expanded role for enoxaparin in preventing venous thromboembolism in high
risk blunt trauma patients. J Am Coll Surg. 2001;192:161-167.
39. Bergqvist D, Lindgren B, Matzsch T. Comparison of the cost of preventing postoperative
deep vein thrombosis with either unfractionated or low molecular weight heparin. Br J
Surg. 1996;83:1548-1552.
40. Etchells E, McLeod RS, Geerts W, Barton P, Detsky AS. Economic analysis of low-dose
heparin vs the low-molecular-weight heparin enoxaparin for prevention of venous
thromboembolism after colorectal surgery. Arch Intern Med. 1999;159:1221-1228.
41. Anderson DR, O'Brien BJ, Levine MN, Roberts R, Wells PS, Hirsh J. Efficacy and cost of
low-molecular-weight heparin compared with standard heparin for the prevention of deep
vein thrombosis after total hip arthroplasty. Ann Intern Med. 1993;119:1105-1112.
42. Hull RD, Raskob GE, Pineo GF, Feldstein W. Subcutaneous low-molecular-weight heparin
vs warfarin for prophylaxis of deep vein thrombosis after hip or knee implantation: an
economic perspective. Arch Intern Med. 1997;157:293-303.
43. Menzin J, Colditz GA, Regan MM, Richner RE, Oster G. Cost-effectiveness of enoxaparin
vs low-dose warfarin in the prevention of deep-vein thrombosis after total hip replacement
surgery. Arch Intern Med. 1995;155:757-764.
44. Wade WE. Cost-effectiveness analysis of deep vein thrombosis prophylaxis in internal
medicine patients. Thrombrosis Research. 1999;94:65-68.
45. Peterson GM, Drake CI, Jupe DM, Vial JH, Wilkinson S. Educational campaign to
improve the prevention of postoperative venous thromboembolism. J Clin Pharm
Therapeutics. 1999;24:279-287.
46. Durieux P, Nizard R, Ravaud P, Mounier N, Lepage E. A clinical decision support system
for prevention of venous thromboembolism. JAMA. 2000;283:2816-2821.
47. Levi D, Kupfter Y, Seneviratne C, Tessler S. Computerized order entry sets and intensive
education improve the rate of prophylaxis for deep vein thrombophlebitis. Chest.
1998;114S:280S.
348

Chapter 32. Prevention of Contrast-Induced Nephropathy
Lorenzo Di Francesco, MD
Mark V. Williams, MD
Emory University School of Medicine
Background
Radiocontrast-induced nephropathy (RCIN) represents an increasingly common cause of
treatment-related renal failure 1-3 and increases mortality independent of other risk factors. 4 Major
risk factors for RCIN include chronic renal insufficiency, 2,3,5 diabetes mellitus 2,3,5 (especially
when accompanied by renal insufficiency 1 ), any condition associated with decreased effective
circulating volume, 6 and use of large doses of contrast media. 2,3,5,6
For at-risk patients, clinicians must use their judgment to determine if imaging modalities
that do not involve contrast media are an acceptable alternative to contrast studies. In many
cases, however, such alternatives do not exist. Moreover, RCIN occurs in patients without
obvious risk factors. Thus, strategies for reducing the incidence of RCIN include not just risk
factor identification, but modification of these risk factors, choice of contrast media less likely to
cause RCIN, and administration of therapeutic agents that further reduce the risk of RCIN.
Practice Description
The specific practices reviewed in this chapter are:
• Use of high versus low osmolar iodinated contrast media to prevent RCIN. 7
• Use of a standard intravenous or oral hydration protocol for patients with risk
factors for RCIN. 8-10 Typical intravenous protocols evaluated consist of normal saline
administered at 75 mL/hr beginning at 12 hours before and ending 12 hours after the
procedure. Oral protocols require ingestion of 1000 mL of water during the 10 hours
prior to the procedure, followed by intravenous normal saline at 300 mL/h for 30-60
minutes and continued for a total of 6 hours after the procedure.
• Use of a standard hydration protocol supplemented by pretreatment with
theophylline 11-15 (various doses and schedules)
• Use of a standard hydration protocol supplemented by pretreatment with Nacetylcysteine
16 (600 mg bid one day before and day of procedure)
Single studies evaluating atrial natriuretic peptide, prostaglandin E1 17 and captopril 18
were not reviewed, as the data are too preliminary, despite findings that suggest a reduction in
the risk of RCIN. Although the evidence supporting the use of N-acetylcysteine largely comes
from a single study as well, we do review this practice because the study was large, published in
a prominent journal, and has received considerable attention among clinicians. 16
The use of calcium channel blockers in preventing RCIN was not evaluated, as the
existing literature predominantly indicates the practice is ineffective. 19-23
349

Prevalence and Severity of the Target Safety Problem
While definitions of RCIN vary, most study definitions include a 25% increase in serum
creatinine (SCr) and/or at least a 0.5 mg/dL increase in SCr within 48 hours of contrast
administration. Using this definition, one large community-based study of 1826 patients
undergoing invasive cardiac procedures reported a rate of RCIN of 14.5%. 2 A controlled
prospective study of the onset of RCIN after contrast-enhanced brain CT found an incidence of
2.1% in low-risk patients without diabetes mellitus or chronic renal insufficiency versus 1.3% in
a similar control group that did not receive any contrast (p=NS). 24 In comparison, patients in a
prospective controlled study undertaken to determine the risk of nephrotoxicity from contrast
radiography in patients with diabetes and renal insufficiency (SCr >1.7mg/dL) found a 9%
incidence of RCIN. 1
The cumulative effect of multiple risk factors increasing the risk of RCIN was
demonstrated in one uncontrolled study that evaluated the effect of 5 factors (contrast volume
>200 mL, albumin

the studies used aminophylline, rather than theophylline). Table 32.1 summarizes the salient
features of these studies.
One randomized trial compared inpatient versus outpatient hydration regimens, 26 but we
found no randomized controlled trial that evaluated pre-hydration versus no hydration. Thus,
support for the standard use of pre-hydration to prevent RCIN is extrapolated from randomized
controlled studies of saline versus saline plus additional pre-treatment agents like mannitol,
furosemide and dopamine 8-10 and smaller observational studies 6,27,28 evaluating the benefits of
pre-hydration.
Study Outcomes
Studies evaluated Level 2 outcomes, primarily by measuring changes in serum creatinine,
creatinine clearance or glomerular filtration, and assessing the frequency of developing acute
renal failure after radiocontrast infusions. Most studies defined RCIN as a 25% increase in
creatinine and/or at least a 0.5 mg/dL increase in serum creatinine within 48 hours of contrast
administration.
Evidence for Effectiveness of the Practice
All of these studies (Table 32.1) evaluated the effects of various prophylactic measures to
reduce the incidence of RCIN. Use of low-osmolar contrast media was supported by one large
meta-analysis 7 that compared low versus high osmolar contrast media. Low osmolar contrast
media was found to be less nephrotoxic than high osmolar contrast media, with an odds ratio for
RCIN of 0.61. Among patients with baseline renal insufficiency (SCr >1.4 mg/dL) the odds ratio
of developing RCIN was 0.5 if low osmolar instead of high osmolar contrast media was used.
As previously noted, no randomized controlled trials have evaluated the efficacy of prehydration
versus no pre-hydration. Data from 3 randomized controlled trials 8-10 using prehydration
versus other pre-treatments and pre-hydration revealed that pre-hydration alone was
equivalent to pre-hydration and low dose dopamine or mannitol, 8 and, in one study, superior to
pre-hydration and furosemide. 10 The incidence of RCIN in patients with SCr >1.6 mg/dL or
creatinine clearance 3 mg/dL, the incidence was only 4%. 8
One retrospective, observational study of high-risk patients undergoing cardiac catheterization
supports the benefit of pre-hydration (>500 mL of 0.9% NS in the pre-catheterization period,
p 0.01) in reducing RCIN. 6 In addition, 2 observational studies without controls 27,28 showed
that pre-hydration in high-risk patients was associated with low rates of RCIN, although one of
these studies 27 used a stricter definition for RCIN (increase in BUN by 50% or 20 mg/dL, and/or
increase in SCr of 1 mg/dL within 24 hours).
A recent study of the oral antioxidant acetylcysteine in combination with pre-hydration in
high-risk patients with renal insufficiency showed significant protective effect against RCIN
versus pre-hydration plus placebo. 16 This protective effect appeared to be even more significant
among patients with more advanced renal dysfunction and SCr >2.5 mg/dL. The overall relative
risk reduction of 90% observed in this study is so large that it raises the possibility of some sort
of bias or other explanation for the observed results. Additional studies of this practice would be
valuable, despite the safety and low cost of N-acetylcysteine.
Studies employing theophylline are more controversial. Three randomized control trials
showed a significant protective effect of various dosages and administration routes of
theophylline among low-risk patients with relatively normal baseline renal function. 12-14 All 3
studies showed theophylline to be protective against a decrease in glomerular filtration rate
351

(GFR) or creatinine clearance (CrCl) after contrast administration. On the other hand, 2 studies
conducted in high-risk patients with renal dysfunction showed no effect for theophylline in
reducing RCIN. 11, 15 Thus, insufficient evidence supports the use of theophylline as prophylaxis
against RCIN in high-risk patients.
Potential for Harm
The impact of a system to identify high-risk patients prior to contrast radiography and
implement aggressive prophylactic measures to reduce the incidence of RCIN has not been
studied. While most patients will not experience any harm from contrast, the potential for
“harm” due to delayed or cancelled investigations may be greater than the harm prevented by
screening for risk factors, aggressive hydration, or use of particular pre-treatment regimens.
Costs and Implementation
At least 4 studies have evaluated the cost-effectiveness of low-osmolality versus highosmolality
contrast media. 29-32 In all 4 studies, the selective use of low-osmolar contrast media
was more cost-effective than its universal use because of the overall small benefits were
outweighed by the considerable increased institutional costs. Alternatively, a standardized
system to identify high-risk patients and implement the simple prophylactic treatment of prehydration
would diminish the frequency of the target problem. It would require collaboration
between the patients’ own physician and the personnel performing the particular contrast study
(radiology department, radiologist, diagnostic/interventional cardiologist). This type of
intervention could be implemented as part of a hospital-based pathway (see Chapter 52) targeted
at reducing radiocontrast-induced nephropathy.
There are no cost-effectiveness or feasibility studies that evaluate protocols for
aggressive identification of high-risk patients undergoing contrast radiography and utilization of
standardized hydration protocols to reduce RCIN. Two studies suggest most patients with
normal renal function (SCr

Table 32.1. Studies of strategies for preventing radiocontrast-induced nephropathy
(RCIN)*
Study Setting Study Design,
Outcomes
Low osmolar contrast media
Meta-analysis of the relative
nephrotoxicity of high (HOCM)
vs. low (LOCM) osmolar
iodinated contrast media 7
Pre-hydration plus diuresis
Patients with SCr >1.8mg/dL
randomized to IVF, IVF +
furosemide, IVF + furosemide +
low dose IV dopamine +/mannitol
(if post-cardiac
catheterization, PCWP 1.6mg/dL or
CrCl 1.7 or CrCl
1.9 mg/dL
who underwent cardiac cath 6
Level 1A,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 3,
Level 2
353
Results
LOCM less nephrotoxic than HOCM; pooled
p=0.02
Odds of ARF with LOCM 0.61 times that of
HOCM (95% CI: 0.48-0.77). Patients with RF at
baseline, odds of ARF were 0.5 (CI: 0.36-0.68).
No differences in rates of renal failure between
groups. Rates of RCIN 21.6% if UOP >150
mL/h, 45.9% if UOP 3 mg/dL, RCIN in patients
with IVF alone 4%, IVF + mannitol 24%
(p=0.02), IVF + furosemide 25% (p=0.02). LOS
increased by 4 days in RCIN group.
SCr increased by 0.42 mg/dL +/- 0.20 treatment
group vs. 0.023 mg/dL +/- 0.073 (p

Table 32.1. Studies of strategies for preventing radiocontrast-induced nephropathy
(cont.)*
Study Setting Study Design,
Outcomes
N-Acetylcysteine
Patients with SCr >1.2 mg/dL or
CrCl 1.5mg/dL
randomized to pre-hydration vs.
pre-hydration with low dose
dopamine or aminophylline prior
to cardiac catheterization. 11
Patients randomized to prehydration
+ theophylline (270 mg
q am/540 mg q pm 2d before, 3d
after) or placebo prior to contrast
CT or DSA. 15
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
Level 1,
Level 2
354
Results
RCIN developed in 2% treatment group vs. 21%
control group (p=0.01). Among patients with
SCr >2.5 mg/dL, RCIN 0% treatment vs. 42%
controls (p=0.02)
GFR reduced 85.4 +/- 3.8 mL/min controls vs.
107 +/-3.6 mL/min treatment group (p 0.001).
CrCl after LOCM decreased by ~18% at 24 hrs
in control (p

References
1. Parfrey PS, Griffiths SM, Barrett BJ, Paul MD, Genge M, Withers J, et al. Contrast
material-induced renal failure in patients with diabetes mellitus, renal insufficiency, or
both. A prospective controlled study. N Engl J Med. 1989;320:143-149.
2. McCullough PA, Wolyn R, Rocher LL, Levin RN, O'Neill WW. Acute renal failure after
coronary intervention: incidence, risk factors, and relationship to mortality. Am J Med.
1997;103:368-375.
3. Rich MW, Crecelius CA. Incidence, risk factors, and clinical course of acute renal
insufficiency after cardiac catheterization in patients 70 years of age or older. A
prospective study. Arch Intern Med. 1990;150:1237-1242.
4. Levy EM, Viscoli CM, Horwitz RI. The effect of acute renal failure on mortality. A cohort
analysis. JAMA. 1996;275:1489-1494.
5. Lautin EM, Freeman NJ, Schoenfeld AH, Bakal CW, Haramati N, Friedman AC, et al.
Radiocontrast-associated renal dysfunction: incidence and risk factors. AJR Am J
Roentgenol. 1991;157:49-58.
6. Brown RS, Ransil B, Clark BA. Prehydration protects against contrast nephropathy in high
risk patients undergoing cardiac catheterization. J Am Soc Neprhol. 1990;1:330A.
7. Barrett BJ, Carlisle EJ. Metaanalysis of the relative nephrotoxicity of high- and lowosmolality
iodinated contrast media. Radiology. 1993;188:171-178.
8. Solomon R, Werner C, Mann D, D'Elia J, Silva P. Effects of saline, mannitol, and
furosemide to prevent acute decreases in renal function induced by radiocontrast agents. N
Engl J Med. 1994;331:1416-1420.
9. Stevens MA, McCullough PA, Tobin KJ, Speck JP, Westveer DC, Guido-Allen DA, et al.
A prospective randomized trial of prevention measures in patients at high risk for contrast
nephropathy: results of the P.R.I.N.C.E. Study. Prevention of Radiocontrast Induced
Nephropathy Clinical Evaluation. J Am Coll Cardiol. 1999;33:403-411.
10. Weinstein JM, Heyman S, Brezis M. Potential deleterious effect of furosemide in
radiocontrast nephropathy. Nephron. 1992;62:413-415.
11. Abizaid AS, Clark CE, Mintz GS, Dosa S, Popma JJ, Pichard AD, et al. Effects of
dopamine and aminophylline on contrast-induced acute renal failure after coronary
angioplasty in patients with preexisting renal insufficiency. Am J Cardiol. 1999;83:260-
263, A5.
12. Kolonko A, Wiecek A, Kokot F. The nonselective adenosine antagonist theophylline does
prevent renal dysfunction induced by radiographic contrast agents. J Nephrol.
1998;11:151-156.
13. Katholi RE, Taylor GJ, McCann WP, Woods WT, Jr., Womack KA, McCoy CD, et al.
Nephrotoxicity from contrast media: attenuation with theophylline. Radiology.
1995;195:17-22.
14. Erley CM, Duda SH, Schlepckow S, Koehler J, Huppert PE, Strohmaier WL, et al.
Adenosine antagonist theophylline prevents the reduction of glomerular filtration rate after
contrast media application. Kidney Int. 1994;45:1425-1431.
15. Erley CM, Duda SH, Rehfuss D, Scholtes B, Bock J, Muller C, et al. Prevention of
radiocontrast-media-induced nephropathy in patients with pre-existing renal insufficiency
by hydration in combination with the adenosine antagonist theophylline. Nephrol Dial
Transplant. 1999;14:1146-1149.
355

16. Tepel M, van der Giet M, Schwarzfeld C, Laufer U, Liermann D, Zidek W. Prevention of
radiographic-contrast-agent-induced reductions in renal function by acetylcysteine. N Engl
J Med. 2000;343:180-184.
17. Koch JA, Plum J, Grabensee B, Modder U. Prostaglandin E1: a new agent for the
prevention of renal dysfunction in high risk patients caused by radiocontrast media? PGE1
Study Group. Nephrol Dial Transplant. 2000;15:43-49.
18. Gupta RK, Kapoor A, Tewari S, Sinha N, Sharma RK. Captopril for prevention of contrastinduced
nephropathy in diabetic patients: a randomised study. Indian Heart J.
1999;51:521-526.
19. Seyss C, Foote EF. Calcium-channel blockers for prophylaxis of radiocontrast-associated
nephrotoxicity. Ann Pharmacother. 1995;29:187-188.
20. Khoury Z, Schlicht JR, Como J, Karschner JK, Shapiro AP, Mook WJ, et al. The effect of
prophylactic nifedipine on renal function in patients administered contrast media.
Pharmacotherapy. 1995;15:59-65.
21. Russo D, Testa A, Della Volpe L, Sansone G. Randomised prospective study on renal
effects of two different contrast media in humans: protective role of a calcium channel
blocker. Nephron. 1990;55:254-257.
22. Neumayer HH, Junge W, Kufner A, Wenning A. Prevention of radiocontrast-mediainduced
nephrotoxicity by the calcium channel blocker nitrendipine: a prospective
randomised clinical trial. Nephrol Dial Transplant. 1989;4:1030-1036.
23. Cacoub P, Deray G, Baumelou A, Jacobs C. No evidence for protective effects of
nifedipine against radiocontrast-induced acute renal failure. Clin Nephrol. 1988;29:215-
216.
24. Cramer BC, Parfrey PS, Hutchinson TA, Baran D, Melanson DM, Ethier RE, et al. Renal
function following infusion of radiologic contrast material. A prospective controlled study.
Arch Intern Med. 1985;145:87-89.
25. Lee JK, Warshauer DM, Bush WH, Jr., McClennan BL, Choyke PL. Determination of
serum creatinine level before intravenous administration of iodinated contrast medium. A
survey. Invest Radiol. 1995;30:700-705.
26. Taylor AJ, Hotchkiss D, Morse RW, McCabe J. PREPARED: Preparation for Angiography
in Renal Dysfunction: a randomized trial of inpatient vs outpatient hydration protocols for
cardiac catheterization in mild-to-moderate renal dysfunction. Chest. 1998;114:1570-1574.
27. Eisenberg RL, Bank WO, Hedgcock MW. Renal failure after major angiography. Am J
Med. 1980;68:43-46.
28. Teruel JL, Marcen R, Herrero JA, Felipe C, Ortuno J. An easy and effective procedure to
prevent radiocontrast agent nephrotoxicity in high-risk patients. Nephron. 1989;51:282.
29. Barrett BJ, Parfrey PS, Foley RN, Detsky AS. An economic analysis of strategies for the
use of contrast media for diagnostic cardiac catheterization. Med Decis Making.
1994;14:325-335.
30. Michalson A, Franken EA, Smith W. Cost-effectiveness and safety of selective use of lowosmolality
contrast media. Acad Radiol. 1994;1:59-62.
31. Caro JJ, Trindade E, McGregor M. The cost-effectiveness of replacing high-osmolality
with low-osmolality contrast media. AJR Am J Roentgenol. 1992;159:869-874.
32. Steinberg EP, Moore RD, Powe NR, Gopalan R, Davidoff AJ, Litt M, et al. Safety and cost
effectiveness of high-osmolality as compared with low-osmolality contrast material in
patients undergoing cardiac angiography. N Engl J Med. 1992;326:425-430.
356

33. Tippins RB, Torres WE, Baumgartner BR, Baumgarten DA. Are screening serum
creatinine levels necessary prior to outpatient CT examinations? Radiology. 2000;216:481-
484.
34. Choyke PL, Cady J, DePollar SL, Austin H. Determination of serum creatinine prior to
iodinated contrast media: is it necessary in all patients? Tech Urol. 1998;4:65-69.
357

358

Chapter 33. Nutritional Support
Neil Winawer, MD
Mark V. Williams, MD
Emory University School of Medicine
Background
There is a consensus that nutritional support should be routinely provided to intensive
care unit (ICU) patients. 1 Hospitalized patients with malnutrition (macronutrient and/or
micronutrient deficiency) suffer from increased infectious morbidity, prolonged hospital stays,
and increased mortality. 2 Moreover, even those hospitalized medical and surgical patients
without antecedent malnutrition are typically subjected to stress, infection and impaired organ
function, resulting in a hypercatabolic state. Often these patients are unable to meet their caloric
needs, as they are either too sick or physically unable to ingest food. Although strong evidence
demonstrates that providing nutritional support for such patients results in improved clinical
outcomes, the optimal method of delivery, timing of administration, and specific formulation
requires further research.
Practice Description
There are several ways to provide nutritional support to patients in the ICU. Enteral
nutrition (EN) can be administered via transoral, transnasal, or percutaneous transgastric routes,
or by surgical jejunostomy. Total parental nutrition (TPN) is generally used when the enteral
route is either inaccessible or its use is contraindicated. It is also used as a supplement to enteral
feeding if adequate nutrition is not possible via the enteral route alone.
The total caloric requirement of critically ill patients can be estimated or directly
measured. Calorimetry, although accurate, is not practical in the clinical setting as it is costly,
time consuming, and requires technical skill. It is also unclear that exactly matching energy input
with energy expenditures improves patient outcomes. Therefore, a pragmatic approach is to
attempt administration of 25 kilocalories per kilogram ideal body weight per day for most
patients. 3 The total caloric daily requirement should be administered in a fluid volume consistent
with the patient’s needs (usually 1mL/kcal). Protein sources should comprise 15-20% of the total
daily calorie requirement. The generally accepted amount of protein is between 1.2 and 1.5 g/kg
per day, except in severe losses such as burns. Glucose should comprise 30-70% of the total
calories and fats 15-30%. 1
Prevalence and Severity of the Target Safety Problem
Malnutrition in hospitalized patients often goes unrecognized. 4,5 Early studies reported a
prevalence of malnutrition in 30-50% of hospitalized patients. 6 A later study revealed that up to
40% of patients were malnourished at the time of their admission. 7 The majority of these patients
continued to be nutritionally depleted throughout their hospital course. These patients are also at
a greater risk for the development of severe malnutrition than those patients whose nutritional
status was adequate at the time of admission.
Unfortunately, there is no single, readily available measure of malnutrition that is both
sensitive and specific in critically ill patients. 1 Most studies have used body mass index
(BMI=weight (kg)/height (m)) and/or anthropometry (measuring skin fold thickness) to assess
patients’ nutritional status. 7 BMI alone is not a sensitive indicator of protein-energy malnutrition
as it does not distinguish between depletion of fat or muscle. 4 In a large number of studies,
359

malnutrition has been defined as a BMI ≤20 kg/m and a triceps skin fold thickness (TSF) or midarm
muscle circumference (MAMC)

Evidence for Effectiveness of the Practice
Nutritional supplementation in hospitalized patients may reduce mortality and is
associated with weight gain (see Table 33.1). 13 However, there are no randomized controlled
trials comparing supplemental nutrition to starvation in critically ill patients. Research does show
that patients not receiving any nutritional support for more than 2 weeks postoperatively have a
much higher complication and mortality rate than patients receiving TPN or some short-term
glucose administration. 14 A large body of research in the past 2 decades has focused on
determining the ideal type and method of delivery of nutritional support.
A meta-analysis comparing supplemental TPN to standard care (oral diet as tolerated and
intravenous dextrose) found no effect on mortality (relative risk (RR) 1.03, 95% CI: 0.81-1.31). 11
There was a trend toward a lower complication rate among those receiving TPN (RR 0.84, 95%
CI: 0.64-1.09), but this is due mainly to benefit among malnourished patients. 11 There are no
data from randomized controlled trials to support the use of supplemental TPN among patients
with an intact gastrointestinal tract (“If the gut works, use it”).
Several studies have evaluated the use of supplemental EN in surgical patients. These
studies often used surrogate outcomes, but one randomized double-blind trial of early EN versus
standard diet as tolerated following surgery found fewer total complications (26.7% vs. 63.3%,
p=0.009), fewer infectious complications (6.7% vs. 46.7%, p

10%) in those patients receiving TPN. 6 This benefit appears to be due entirely to significant
reduction in surgical morbidity among patients who are severely malnourished. 22 Therefore,
preoperative TPN may be of benefit in severely malnourished patients undergoing major
gastrointestinal surgery, but EN should be used instead, if possible. The use of postoperative
TPN has been evaluated in 8 prospective randomized trials of patients undergoing
gastrointestinal surgery. Patients in the combined TPN group experienced an increased rate of
complications (27.3% vs. 16.0%; p

developing major infectious complications (odss ratio 0.47, 95% CI: 0.32-0.70), and hospital
length of stay (reduced 2.5 days, 95% CI: 1.0-4.0). 29 Though there was no “significant”
difference in mortality between patients receiving immune-enhanced versus standard nutritional
support, there was a trend toward increased mortality (odds ratio 1.77, 95% CI: 1.00-3.12). 29
Potential for Harm
TPN has been associated with an increase in septic complications. Patients may also be
placed at increased risk from attempts to obtain central venous access. While EN decreases the
overall rate of infectious complications when compared to TPN, it may place patients, especially
those with head injuries, at greater risk of aspiration pneumonia. Use of a promotility agent, such
as metoclopramide, has reduced this risk in one study. 30
Costs and Implementation
TPN is itself expensive. Moreover, hospital costs can climb dramatically if patients who
receive TPN suffer infectious complications that prolong their hospital stay. Obtaining central
venous access also increases the costs of administration. Nutritional support teams (NSTs) can
help ensure that patients are receiving adequate nutrition while at the same time reducing the rate
of line complications. 28 However, the use of NSTs is expensive and has not been subject to a
cost analysis. 28 The economics of early postoperative enteral nutrition was studied in a recent
nonrandomized, prospective, clinical trial of patients undergoing bowel resections. There was a
reduction in variable cost of $1531 per success in the treatment group (p=0.02) and $4450 total
cost savings per success in the treatment group (p=0.04). 31 Immunonutrition is significantly more
expensive than standard enteral formulas but in 2 prospective randomized controlled trials of
patients undergoing upper gastrointestinal surgery it has been shown to be cost-effective. In one
study, the cost per patient for nutrition and complications in the control group was $1633
compared to $783 in the immunonutrition group. 32 In another study the cost per patient for
nutrition and complications in the control group was $1107 compared with $755 in the treatment
group. 33
Comment
The literature on appropriate nutritional support for critically ill patients is complex to
analyze. A wide variety of treatment modalities have been evaluated in a fairly heterogeneous
patient population, making interpretation of the various studies difficult.
Nonetheless, there are several general recommendations that can be made. First,
malnutrition leads to poor outcomes and should be avoided or treated if present. Patients who
have prolonged starvation for more than 2 weeks are at a significantly increased risk for
complications. EN in these patients decreases this risk, and should be administered to all
critically ill medical and surgical patients who can tolerate it. There are no data to support the
use of TPN in critically ill patients who have a functional gastrointestinal tract. However, TPN
may be of benefit preoperatively when given electively to severely malnourished patients
undergoing gastrointestinal resection. Postoperative administration of TPN worsens outcomes.
There is no evidence to suggest that postoperative enteral nutrition is superior to advancing a
patient’s oral diet, as tolerated, if they are capable of eating, but research into early
administration of EN via the jejunum is ongoing.
Use of immune-enhanced EN appears to significantly reduce infectious complications,
number of days of mechanical ventilation, and overall hospital length of stay, though it does not
appear to affect overall mortality (and may increase it). Of note, study groups with increased
363

mortality rates may also have a diminished incidence of infectious complications and decreased
length of stay artifactually due to patients dying early in the study. One expert interpreted the
published research as strongly suggesting that immune-enhancing formulas should be used in
critically ill patients, especially surgical patients, despite their high cost. 34 Nonetheless, a large
multicenter RCT is needed to resolve with certainty whether or not use of these expensive
formulas is safe and improves outcomes in critically ill patients before formally recommending
their widespread use. 29
364

Table 33.1. Studies evaluating nutritional support*
Study Study Design,
Outcomes
Routine protein energy
supplementation in
adults; systematic
review 13
TPN in critically ill
patients; metaanalysis
11
TPN in surgical
patients; metaanalysis
12
Immunonutrition in
critically ill patients;
systematic review and
meta-analysis 28
Immunonutrition in
patients with critical
illness and cancer;
meta-analysis 29
EN vs. TPN in
seriously ill
hospitalized patients;
systematic review and
meta-analysis 2
Enteral
immunonutrition in the
critically ill; doubleblind
randomized
controlled trial 26
Level 1A, Level
1
Level 1A, Level
1
Level 1A, Level
1
Level 1A, Level
1
Level 1A, Level
1
Level 1A, Level
1
Level 1, Level
1
Results (95% Confidence Interval)
Reduced mortality: OR 0.66 (0.48 to 0.91)
Increased body weight gain (%):
Oral Sip Feeds: 2.39 (1.80 to 2.96)
Oral natural feeds: 5.36 (1.73 to 8.99)
Nasogastric feeding: 4.04 (3.15 to 4.94)
Percutaneous or enteral feeding, enterostomy:
-1.38 (-2.35 to -0.41)
No effect on mortality: RR 1.03 (0.81 to 1.31)
Complication rate: RR 0.84 (0.64 to 1.09)
Complication rate (malnourished subgroup):
RR 0.52 (0.30 to 0.91)
No effect on mortality: RR 0.97 (0.76 to 1.24)
No effect on mortality (malnourished subgroup):
RR 1.13 (0.75 to 1.71)
Major complication rate (malnourished subgroup):
RR 0.52 (0.30 to 0.91)
No effect on mortality: RR 1.05 (0.78 to 1.41)
Reduction in infection rate: RR 0.67 (0.50 to 0.89)
Reduction in days of mechanical ventilation:
2.6 days (0.1 to 5.1)
Reduction in hospital LOS: 2.9 days (1.4 to 4.4)
Trend toward increased mortality:
OR 1.77 (1.00 to 3.12)
Reduction in infectious complications:
OR 0.47 (0.32 to 0.70)
No reduction in pneumonia risk: OR 0.91 (0.53 to 1.56)
Reduction in hospital LOS: 2.5 days (4.0 to 1.0)
Reduction in septic complications:
EN 18% vs. TPN 35% (p=0.01)
Reduction in infectious complications:
EN 16% vs. TPN 35% (p=0.01)
Overall analysis (immunonutrition vs. control formula)
Mortality: 48% vs. 44% (p=0.36)
Total days of mechanical ventilation (median):
4 vs. 4 days (p=NS)
Hospital LOS (median): 12 vs. 13 days (p=NS)
Early enteral feeding subgroup
Total days of mechanical ventilation (median):
6 vs. 10.5 days (p=0.007)
Hospital LOS (median): 15.5 vs. 20 days (p=0.03)
* LOS indicates length of stay; NS, not statistically significant; OR, odds ratio; and RR, relative risk.
365

References
1. Cerra FB, Benitez MR, Blackburn GL, Irwin RS, Jeejeebhoy K, Katz DP, et al. Applied
nutrition in ICU patients. A consensus statement of the American College of Chest
Physicians. Chest. 1997;111:769-778.
2. Heyland DK. Enteral and parenteral nutrition in the seriously ill, hospitalized patient: A
critical review of the evidence. Journ of Nutr Health & Aging. 2000;1:31-41.
3. Jolliet P, Pichard C, Biolo G, Chiolero R, Grimble G, Leverve X, et al. Enteral nutrition in
intensive care patients: a practical approach. Intensive Care Med. 1998;24:848-859.
4. McWhirter JP, Pennington CR. The incidence and recognition of malnutrition in hospital.
BMJ. 1994;308:945-948.
5. Potter JM, Klipstein K, Reilly JJ, Roberts MA. The nutritional status and clinical course of
acute admissions to a geriatric unit. Age Ageing. 1995;24:131-136.
6. Torosian MH. Perioperative nutrition support for patients undergoing gastrointestinal
surgery: critical analysis and recommendations. World J Surg. 1999;23:565-569.
7. Beattie AH, Prach AT, Baxter JP, Pennington CR. A randomized controlled trial evaluating
the use of enteral nutritional supplements postoperatively in malnourished surgical patients.
Gut. 2000;46:813-818.
8. Adam S, Batson S. A study of problems associated with the delivery of enteral feed in
critically ill patients in five ICUs in the UK. Intensive Care Med. 1997;23:261-266.
9. McClave SA, Sexton LK, Spain DA. Enteral tube feeding in the intensive care unit: factors
impeding adequate delivery. Crit Care Med. 1999;27:1252-1256.
10. DeJonghe B, Appere-De-Vechi C, Fournier M, Tran B, Merrer J, Melchior JC, et al. A
prospective survey of nutritional support practices in intensive care unit patients: What is
prescribed? What is delivered? Crit Care Med. 2001;29:8-12.
11. Heyland DK, MacDonald S, Keefe L, Drover JW. Total parental nutrition in the critically
ill patient. A meta-analysis. JAMA. 1998;280:2013-2019.
12. Heyland DK, Montalvo M, MacDonald S, Keefe L, Su XY, Drover JW. Total parenteral
nutrition in the surgical patient: a meta-analysis. Can J Surg. 2001;44:102-111.
13. Potter J, Langhorne P, Roberts M. Routine protein energy supplementation in adults:
systematic review. BMJ. 1998;317:495-501.
14. Sandstrom R, Drott C, Hyltander A. The effect of postoperative intravenous feeding (TPN)
on outcome following major surgery evaluated in a randomized study. Ann Surg.
1993;217:185-195.
15. Beier-Holgersen R, Boesby S. Influence of postoperative enteral nutrition on postsurgical
infections. Gut. 1996;39:833-835.
16. Singh G, Ram RP, Khanna SK. Early postoperative enteral feeding in patients with
nontraumatic intestinal perforation and peritonitis. J Am Coll Surg. 1998;187:142-146.
17. Kalfarentzos F, Kehagias J, MEad N, Kokkinis K, Gogos CA. Enteral nutrition is superior
to parenteral nutrition in severe acute pancreatitis: Results of a randomized prospective
trial. Br J Surg. 1997;84:1665-1669.
18. Iovinelli G, Marsili I, Varassi. Nutritional support after total laryngectomy. JPEN.
1993;17:445-448.
366

19. Moore FA, Moore EE, Jones TN, McCroskey BL, Petersen VM. TEN versus TPN
following major abdominal trauma reduced septic morbidity. J of Trauma. 1989;29:916-
923.
20. Moore FA, Feliciano DV, Andrassy RJ, McArdle AH, Booth FV, Morgenstein-Wagner
TB, et al. Early enteral feeding, compared with parental, reduces postoperative septic
complications. The results of a meta-analysis. Ann Surg. 1992;216:172-183.
21. Young B, Ott L, Twyman. The effect of nutritional support on outcome from severe head
injury. J Neurosurg. 1987;67:668-676.
22. The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. Perioperative
total parenteral nutrition in surgical patients. N Engl J Med. 1991;325:525-532.
23. Zaloga GP, Bortenschlager L, Black KW, Prielipp R. Immediate postoperative enteral
feeding decreases weight loss and improves healing after abdominal surgery in rats. Crit
Care Med. 1992;20:115-119.
24. Mochizuki H, Trocki O, Dominioni L, Brackett KA, Joffe SN, Alexander JW. Mechanism
of prevention of postburn hypermetabolism and catabolism by early enteral feeding. Ann
Surg. 1984;200:297-310.
25. Moore EE, Jones TN. Benefits of immediate jejunostomy feeding after major abdominal
trauma - a prospetive randomized study. J of Trauma. 1986;26:874-880.
26. Atkinson S, Sieffert E, Bihari D. A prospective, randomized, double-blind, controlled
clinical trial of enteral immunonutrition in the critically ill. Crit Care Med. 1998;26:1164-
1172.
27. Heyland DK, Drover J. Does immunonutrition make an impact? It depends on the analysis.
Crit Care Med. 2000;28:906.
28. Beale RJ, Bryg DJ, Bihari DJ. Immunonutrition in the critically ill: A systematic review of
clinical outcome. Crit Care Med. 1999;27:2799-2805.
29. Heys SD, Walker LG, Smith I, Eremin O. Enteral nutritional supplementation with key
nutrients in patients with critical illness and cancer. A meta-analyis of randomized
controlled clinical trials. Ann Surg. 1999;229:467-477.
30. Jooste CA, Mustoe J, Collee G. Metoclopraamide improves gastric motility in critically ill
patients. Intensive Care Med. 1999;25:464-468.
31. Hedberg A, Lairson DR, Aday LA, Chow J, Suki R, Houston S, et al. Economic
implications of an early postoperative enteral feeding protocol. J Am Diet Assoc.
1999;99:802-807.
32. Senkal M, Zumtobel V, Bauer KH, Marpe B, Wolfram G, Frei A, et al. Outcome and costeffectiveness
of perioperative enteral immunonutrition in patients undergoing elective
upper gastrointestinal tract surgery. Arch Surg. 1994;134:1309-1316.
33. Senkal M, Mumme A, Eickhoff U, Geier B, Spath G, Wulfert D, et al. Early postoperative
enteral immunonutrition: Clinical outcome and cost-comparison analysis in surgical
patients. Crit Care Med. 1997;25:1489-1496.
34. Zaloga GP. Immune-enhancing diets: Where's the beef? Crit Care Med. 1998;26:1143-
1146.
367

Chapter 34. Prevention of Clinically Significant Gastrointestinal Bleeding in
Intensive Care Unit Patients
Daniel D. Dressler, MD
Mark V. Williams, MD
Kimberly Rask, MD, PhD
Emory University Schools of Medicine and Public Health
Background
Stress-related gastric ulceration was first described in the early 1970s, 1 and has since
received extensive study. Appreciation of the physiologic changes that promote stress-related
gastritis 2-4 and general improvements in the care of critically ill patients have likely played a role
in reducing the frequency of this complication. 3 Nonetheless, the use of specific pharmacologic
agents for the prevention of stress-related gastrointestinal (GI) bleeding is increasingly promoted
as standard therapy in the ICU setting. 5-8 Despite the common use of these agents, controversy
about the evidence supporting this practice remains. 5,9 Because specific pharmacologic
prophylaxis for stress ulceration may increase the risk of other complications (eg, nosocomial
pneumonia 10 ) and is associated with significant costs, 11 recent efforts have focused on
delineating an appropriate definition of clinically important GI bleeding and identifying patients
who derive a clear benefit from pharmacologic prevention of this complication.
Practice Description
This chapter reviews evidence supporting the use of pharmacologic therapies for stress
ulcer and GI bleeding prophylaxis in the ICU setting. We considered the use of histamine-2
receptor blocking agents (H2-receptor antagonists) and sucralfate, a mucosal protecting agent.
Although the efficacy of aluminum- and magnesium-based antacids has been demonstrated, their
use is limited by dosing frequency and side effects. 12 Proton-pump inhibitors (PPIs) and enteral
nutrition have also shown benefit in small studies, but there are yet no large randomized
evaluations. 6,12,13
Prevalence and Severity of the Target Safety Problem
The risk of stress ulceration and subsequent GI bleeding depends on a patient’s
underlying illness, its severity, and related comorbidities. Using liberal criteria for assessment of
bleeding, early reports estimated the incidence of GI bleeding due to stress ulceration in ICU
patients to be as high as 5-25%. 14,15 However, more recent prospective studies, using stricter
definitions of clinically significant GI bleeding and following large cohorts of patients, have
revealed more modest estimates of 0.1% in low-risk ICU patients and 2.8% in ventilated
patients. 2,13,16
Opportunities for Impact
A recent, prospectively collected database of over 7000 patients admitted to surgical and
medical intensive care units at a tertiary care center (1988-95) revealed that 59.9% of patients
received pharmacologic prophylaxis with ranitidine or sucralfate for stress ulceration. 16 Another
study assessed the annual change in stress ulcer prophylaxis in nearly 3000 patients admitted to a
medical intensive care unit at a tertiary care hospital from 1993 to 1996. It found a significant
decrease in rates of prophylaxis, with a 71% rate in 1993 progressively decreasing to a 21% rate
in 1996, likely reflecting emerging evidence regarding high and low-risk populations in the
368

literature during that time. 17 Identifying appropriate patients for stress ulcer prophylaxis and
avoiding its use in other cases both offer potential opportunities for improving patient safety in
the ICU.
Study Designs
Although multiple meta-analyses have reviewed the prophylaxis of GI bleeding in
critically ill patients, many were performed more than a decade ago. 18-20 Since then, results of
additional randomized trials have been published. More recent, well-designed search strategies,
strict quality-scoring of literature, and rigorous adherence to criteria defining clinically
significant GI bleeding have improved the breadth and quality of literature captured, likely
making older systematic reviews less relevant. Because of these changes in quality measures as
well as overlap of information and trials from older reviews contained within more recent ones,
only those meta-analyses published within the last 5 years (and large randomized controlled
trials not included in these reviews) are included here.
A 1996 meta-analysis found 269 studies through a broad search of multiple databases and
extensive efforts to find unpublished trials. 5 Sixty-three of these met inclusion criteria. The
analysis evaluated multiple types of prophylaxis for GI bleeding, including antacids, H2antagonists,
and sucralfate. The study authors compared each type of prophylaxis to placebo and
to each other when literature was available. The same study group then followed-up unanswered
questions from their meta-analysis with a large randomized controlled trial (RCT) of 1200
patients, comparing ranitidine to sucralfate without a placebo group. 21
A more recent meta-analysis abstracted evidence from the literature to compare ranitidine
(instead of all H2-antagonists) versus placebo, and sucralfate versus placebo for GI bleed
prophylaxis in the ICU. 9 The search methods were less rigorous and the target population
narrower than the 1996 meta-analysis, but the authors claimed to assess a more clinically
relevant population and outcome.
Results from 2 prospective cohort studies that evaluated risk factors for occurrence of
clinically important GI bleeding also provide key recommendations regarding clinical
practice. 2,13
Study Outcomes
Both meta-analyses and the large RCT reported outcomes of clinically important GI
bleeding and nosocomial pneumonia (Level 1). The 1996 meta-analysis and randomized trial
also evaluated overt GI bleeding and mortality. Cohort studies (Level 2) present results of
multivariate regression analyses, identifying the salient risk factors for clinically important GI
bleeding in the ICU patient.
Evidence for Effectiveness of the Practice
Table 34.1 summarizes the outcomes for the 2 meta-analyses and large RCT. The
disparity in results of the meta-analyses is partly due to differing article selection criteria. While
the 1996 analysis evaluated all H2-antagonists, the 2000 study excluded trials evaluating
cimetidine because of its virtual replacement (attributable to its undesirable side effect profile)
by ranitidine in current clinical practice. 5,9 Excluding studies with cimetidine substantially
reduced the number of studies available for review. Additionally, the 1996 study evaluated
comparisons of therapies to each other as well as to placebo, while the 2000 study only
considered trials of effectiveness with a placebo control as relevant.
369

The 2000 meta-analysis found no statistically significant reduction in clinically important
GI bleeding when comparing H2-antagonist agents with placebo, or sucralfate to placebo. 9
However, the 1996 study revealed a significant difference between ranitidine and placebo for
clinically important GI bleeding. 5 It was unable to demonstrate such a difference between
ranitidine and sucralfate. The results of these 2 reviews reveal discrepant findings, in large part
due to exclusion of cimetidine in the 2000 study. However, that study also included at least one
moderate-sized trial that used remarkably strict criteria for defining clinically important GI
bleeding. This factor likely contributed an element of bias towards the null result in that metaanalysis.
Furthermore, the large RCT (1998) comparing ranitidine to sucralfate, without a placebo
comparison group, revealed a statistically significant difference in the rate of clinically important
upper GI bleeding, favoring H2-antagonists. 21 In this RCT the number needed to treat (NNT)
with ranitidine compared to placebo to prevent one clinically important GI bleed compared to
placebo was 47. However, no reduction in mortality or length of ICU stay was found.
Interpretation of the results of the RCT are complicated by: 1) wide confidence intervals
surrounding the relative risk estimate; 2) very small numbers of patients with clinically
important GI bleeding (ranitidine group: 10/596; sucralfate group: 23/604), only 42% of whom
had endoscopic confirmation of the source of bleed; 3) a large number of patients (70%)
receiving concomitant enteral nutrition, believed to reduce the risk of GI bleeding (see Chapter
33); and 4) unreported incidence of coagulopathy or duration of prophylaxis prior to GI
bleeding. 22
The 2 large cohort studies found respiratory failure (odds ratio 15.6, p

Costs and Implementation
One cost-effectiveness analysis extracted relevant literature from MEDLINE between
1985 and 1995, representing 15 controlled clinical trials (Level 1). 19 Assumptions made in the
analysis included equal efficacy for the H2-antagonist cimetidine and sucralfate, a baseline risk
of bleeding of 6%, and a 50% risk-reduction due to prophylaxis. The average cost per bleeding
episode averted was $1144 for sucralfate, but 6.5 times greater for cimetidine. However, low-risk
patients amassed a cost per bleeding episode averted of $103,725 compared to a cost of $279 for
very high-risk patients. Cost per bleeding episode averted increased significantly if the risk of
nosocomial pneumonia was included in the analysis.
The large RCT reported no difference in the duration of ICU stay for the group treated
with ranitidine compared with the group treated with sucralfate (9 days for each group). 21
However, no placebo group was used.
Comment
Overall, the evidence available in the literature does not conclusively demonstrate that
the benefits of GI prophylaxis outweigh its risks for every patient admitted to the intensive care
unit. As refinements have been made in defining clinically important and significant episodes of
GI bleeding, the population that may reap benefit from prophylaxis has narrowed substantially.
In turn, a reduction in the use of prophylactic agents has followed. Most recent estimates reveal a
negligible incidence of clinically important stress-related GI bleeding in low-risk ICU patients
(approximately 0.1%) and a small incidence (less than 3%) in higher-risk patients. Improvements
in overall ICU care and use of enteral nutrition may be contributing to this decrease in incidence.
Although a statistical benefit of H2-antagonists compared with placebo or with sucralfate
has been shown in some studies, the overall clinical benefit of these agents has been disputed.
Some research shows a greater absolute number of nosocomial pneumonia cases related to
therapy with H2-antagonists than the benefit from reduction in GI bleeding, causing the NNH to
be potentially smaller than the NNT (see Subchapter 17.4). Furthermore, the overall cost-tobenefit
ratio of prophylaxis increases dramatically in lower-risk patients, as shown in the
analysis noted above that accounted for increases in nosocomial pneumonia. 21 Thus, the clear
benefit of administering H2-antagonists to prevent GI bleeding among many patients in an ICU
remains to be shown. This is partly due to variations in the definition of clinically significant
bleeding and pneumonia. Additional large trials are necessary to identify the patients who truly
derive net benefit from this practice, as published evidence does not yet support this practice for
many patients currently receiving the therapy.
At the present time, clinicians may consider use of prophylactic agents, an H2-antagonist
or sucralfate, to prevent clinically important GI bleeding in very high-risk patients admitted to
the ICU. Such patients may include those with respiratory failure, coagulopathy, renal failure,
and/or burns (the latter group has been excluded from most studies because its risk is believed to
be so great). However, the risk of pneumonia may influence clinicians to use prophylactic agents
only in patients with multiple risk factors for GI bleeding, and simply provide enteral nutrition to
others at less risk. The use of PPIs requires further study before any recommendation can be
made regarding them. Further research should focus on identifying subgroups of patients who do
derive net benefit from ranitidine or other acid-reducing agents.
371

Table 34.1. Studies evaluating effectiveness of pharmacologic prophylaxis of ICU patients
to prevent clinically significant GI bleeding*
EFFECTIVENESS
Study Study Design, Outcomes Comparison Groups Effect Size (95% CI)
Cook,
1996 5
Messori,
2000 9
Cook,
1998 21
Level 1A, Level 1 H2-antagonist vs. placebo
Sucralfate vs. placebo
H2-antagonist vs. sucralfate
372
OR 0.44 (0.22-0.88)
OR 1.49 (0.42-5.27)
OR 1.28 (0.27-6.11)
Level 1A, Level 1 Ranitidine vs. placebo OR 0.72 (0.30-1.70)
Level 1, Level 1 Ranitidine vs. sucralfate RR 0.44 (0.21-0.92)
NNT 47
* CI indicates confidence interval; NNT, number needed to treat (for benefit); OR, odds ratio;
and RR, relative risk.
Table 34.2. Studies evaluating harm (nosocomial pneumonia) due to pharmacologic
prophylaxis of ICU patients to prevent GI bleeding*
Study Study Design, Outcomes
Measured
Cook,
1996 5
Messori,
2000 9
Cook,
1998 21
HARM
Comparison Groups Effect Size (95% CI)
Level 1A, Level 1 Sucralfate vs. H2-antagonist OR 0.78 (0.60-1.01)
Level 1A, Level 1 Ranitidine vs. sucralfate OR 1.35 (1.07-1.70)
NNH 21
Level 1, Level 1 Ranitidine vs. sucralfate RR 1.18 (0.92-1.51)
NNH 34
* CI indicates confidence interval; NNH, number needed to harm; OR, odds ratio; and RR,
relative risk.
References
1. Lucas C, Sugawa C, Riddle J, Rector F, Rosenberg B, Walt A. Natural History and surgical
dilemma of "stress" gastric bleeding. Arch Surgery. 1971;102:266-273.
2. Cook D, Fuller H, Guyatt G, Marshall J, Leasa D, Hall R, et al. Risk factors for
gastrointestinal bleeding in critically ill patients. N Engl J Med. 1994;330:377-381.
3. Haglund U. Stress ulcers. Scand J Gastroenterol. 1990;175:27-33.
4. Shoemaker W, Bland R, PL A. Therapy of critically ill postoperative patients based on
outcome prediction and prospective clinical trials. Surg Clin North Am. 1985;65:811-833.
5. Cook D, Reeve B, Guyatt G, Heyland D, Griffith L, Buckingham L, et al. Stress ulcer
prophylaxis in critically ill patients: resolving discordant meta-analyses. JAMA.
1996;275:308-314.

6. Beejay U, Wolfe M. Acute gastroinestinal bleeding in the intensive care unit: the
gastroenterologist's perspective. Gastroenterol Clin North Am. 2000;29:309-336.
7. Friedman S, Manaker S, Weinhouse G. Stress ulcer prophylaxis in the intensive care unit.
In: Rose BD, ed. UpToDate. Wellesley, MA: UpToDate, Inc.; 2001.
8. ASHP Commission on Therapeutics. Therapeutic guidelines on stress ulcer prophylaxis.
Am J Health Syst Pharm. 1999;56:347-379.
9. Messori A, Trippoli S, Vaiani M, Gorini M, Corrado A. Bleeding and pneumonia in
intensive care patients given ranitidine and sucralfate for prevention of stress ulcer: metaanalysis
of randomised controlled trials. BMJ. 2000; 21:1-7.
10. Ben-Menachem T, McCarthy B, Fogel R, Schiffman R, Patel R, Zarowitz B, et al.
Prophylaxis for stress-related gastrointestinal hemmorhage: a cost effectiveness analysis.
Crit Care Med. 1996;24:338-345.
11. Cook D. Stress ulcer prophylaxis gastrointestinal bleeding and nosocomial pneumonia.
Best evidence synthesis. Scand J Gastroenterol. 1995;210:48-52.
12. Hastings P, Skillman J, Bushnell L, Silen W. Antacid titration in the prevention of acute
gastrointestinal bleeding: a controlled, randomized trial in 100 critically ill patients. N Engl
J Med. 1978;298:1041-1045.
13. Raff T, Germann G, Hartmann B. The value of early enteral nutrition in the prophylaxis of
stress ulceration in the severely burned patient. Burns. 1997;23:313-318.
14. Ben-menachem T, Fogel R, Patel R, Touchette M, Zarowitz B, Hadzijahic N, et al.
Prophylaxis for stress-related gastric hemorrhage in the medical intensive care unit: a
randomized, controlled, single-blind study. Ann Intern Med. 1994; 21:568-575.
15. Shuman R, Schuster D, Zuckerman G. Prophylactic therapy for stress ulcer bleeding: a
reappraisal. Ann Int Med. 1987;106:562-567.
16. Pimentel M, Roberts D, Bernstein C, Hoppensack M, Duerksen D. Clinically significant
gastrointestinal bleeding in critically ill patients in an era of prophylaxis. Am J
Gastroenterol. 2000;95:2801-2806.
17. Devlin J, Ben-Menachem T, Ulep S, Peters M, Fogel R, Zarowitz B. Stress ulcer
prophylaxis in medical ICU Patients: annual utilization in relation to the incidence of
endoscopically proven stress ulceration. Ann Pharmacother. 1998;32:869-874.
18. Tryba M. Prophylaxis of stress ulcer bleeding: a meta-analysis. J Clin Gastroenterol.
1991;3(Suppl. 2):S44-S55.
19. Lacroix J, Infante-Rivard C, Jenicek M, Gauthier M. Prophylaxis of upper gastrointestinal
bleeding in intensive care units: a meta analysis. Crit Care Med 1989;17:862-869.
20. Cook D, Witt L, Cook R, Guyatt G. Stress ulcer prophylaxis in the critically ill: a meta
analysis. Am J Med. 1991;91:519-527.
21. Cook D, Guyatt G, Marshall J, Leasa D, Fuller H, Hall R, et al. A comparison of sucralfate
and ranitidine for the prevention of upper gastrointestinal bleeding in patients requiring
mechanical ventilation. N Engl J Med. 1998;338:791-797.
22. Hinds C, Fletcher S. Ranitidine reduced clinically important gastrointestinal bleeding in
patients who required mechanical ventilation. Gut. 1999;44:10-11.
23. Burgess P, Larson G, Davidson P, Brown J, Metz C. Effect of ranitidine on intragastric pH
and stress-related upper gastrointestinal bleeding in patients with severe head injury. Dig
Dis Sci. 1995;40:645-650.
24. Cook D, Heyland D, Griffith L, Cook R, Marshall J, Pagliarello J. Risk factors for
clinically important upper gastrointestinal bleeding in patients requiring mechanical
ventilation. Crit Care Med. 1999;27:2812-2817.
373

374

375

Chapter 35. Reducing Errors in the Interpretation of Plain Radiographs and
Computed Tomography Scans
Sunil Kripalani, MD
Mark V. Williams, MD
Kimberly Rask, MD, PhD
Emory University Schools of Medicine and Public Health
Background
Misinterpretation of radiographic studies is a common source of medical error in both the
inpatient and outpatient arenas. 1,2 Of particular concern is the significant number of
misinterpretations of plain radiographs and cranial computed tomography (CT) scans in
emergency departments (ED) or urgent care settings by non-radiologists. 3-6 The prevalence of
this patient safety issue may result from the large volume of patients receiving these radiologic
tests, which are often done outside normal working hours when radiologists are not available to
provide an initial interpretation. This chapter focuses on practices to reduce non-radiologists’
higher rates of misinterpretation of these commonly ordered studies.
Intuitively, it would seems that institutions could minimize the number of such mistakes
by routinely having studies interpreted by the most accurate and experienced physicians, usually
radiologists. The American College of Radiology (ACR) recommends that all imaging
procedures culminate in a written, expert opinion from a radiologist or another licensed
physician specifically trained in diagnostic radiology. 7,8 However, due to the associated costs,
fewer than 20% of hospitals have full-time on-site coverage by a board-certified radiologist. 9
Instead, radiologists are generally available for 8 to 12 hours a day and may not provide an
interpretation until the following morning, particularly when studies are performed after normal
working hours. 9 For many routine examinations, it may be possible to delay interpretation
without harm to patients. If results are needed urgently and a radiologist is unavailable, other
physicians (eg, emergency physicians, hospitalists, neurologists) must take responsibility for the
initial interpretation.
Patient safety may be enhanced by improving the diagnostic accuracy of these physicians
or by implementing other systems to prevent initial misinterpretations from adversely affecting
patient care. Several strategies reviewed here that may be effective in reducing these errors
include educational courses to improve the diagnostic accuracy of non-radiologists, on-site
coverage by radiology residents, and mandatory subsequent reinterpretation of studies by
radiologists.
Another approach that deserves mention is to have the initial readings made by off-site
radiologists or other specialists using a teleradiology link. Teleradiology has been effective in
facilitating emergent neurosurgical consultation prior to the interhospital transfer of patients with
head injury. 10-12 Teleradiology also allows rural physicians to obtain remote consults for selected
patients, 13-16 which in one report led to treatment changes in 26% of cases. 17 Despite
teleradiology’s impact in these two specific settings, few studies have tested its accuracy and
utility in more general circumstances. 18 Teleradiology of course requires the use of digitized
rather than film-based radiographs. In two mixed case series, discrepancies of interpretation
between digitized and original radiographs occurred in approximately 10% of cases, 13,14 with
significant discrepancies in 1.5-5%. 19,20 For more subtle findings the sensitivity of on-screen
images may be as low as 49%, 21 leading several investigators to conclude that teleradiology is
376

inferior to film interpretation for difficult cases. 21-23 At present, it appears that image quality is
the major reason behind the variable performance of teleradiology. Although the ACR has
established detailed standards for equipment and image resolution, 24 radiology practices often
utilize less expensive alternatives for viewing images, including personal computers. 23,25,26 The
variation in practice, rapid evolution of technology, and lack of large prospective trials make it
difficult to examine teleradiology from the standpoint of patient safety. Therefore, this chapter
will not discuss teleradiology in detail; reviews can be found elsewhere. 27-30
Interventions to improve the technical quality of imaging studies, such as ensuring proper
patient positioning and film exposure, are certainly important but are also outside the scope of
this chapter. Finally, although radiologists make mistakes in interpreting films, 31-37 this chapter
focuses on practices to reduce the higher rate of misinterpretations made by non-radiologists.
Practice Descriptions
Training Courses
Training courses for non-radiologists are presumably common, but we could locate only
2 descriptions in the literature. 38,39 Both courses concerned interpretation of cranial CT scans,
were conducted by radiologists, and targeted ED residents and faculty physicians. The training
lasted 1-2 hours and consisted of a review of neuroanatomy and a small library of CT scans. Two
continuous quality improvement initiatives were also recently described, in which radiologists
provided regular feedback to ED physicians about their radiograph interpretations. 40,41
Initial Interpretations by Radiology Residents
Initial interpretation of ED films by radiology residents follows no standardized practice.
Their hours of coverage and the degree to which they interact with ED physicians vary widely. 42
It is also unclear to what extent emergency physicians rely on the resident interpretations when
they are available.
Review of All Studies by Radiologists
Reinterpretation of all studies by a radiologist is already commonly used. When a
radiologist finds that a study was initially misinterpreted, the medical error has already occurred.
In this case, the safety practice concerns how the radiologist communicates the corrected
interpretation to providers in order to minimize the risk of harm to patients. The method of
communication varies among health care facilities and includes placing the interpretation in the
patient’s medical record, sending a report to the referring physician, or contacting the physician
or patient directly for more urgent concerns (see also Subchapter 42.4). 7
Prevalence and Severity of the Target Safety Problem
Many investigators have focused on the prevalence of all ED readings that are discordant
with radiologists’ subsequent interpretations (with the radiologists’ interpretation assumed to be
the “gold standard”). Against this standard, ED physicians and residents misinterpret 1-16% of
plain radiographs 43-51 and approximately 35% of cranial CT scans. 52 However, many discordant
readings involve subtle or incidental findings that are not clinically significant (ie, they do not
affect patient management or outcome). From the standpoint of patient safety, it is the rate of
clinically significant misinterpretations and related errors in management that are of concern.
Most ED studies show that important errors in interpretation occur in 1-3% of plain
radiographs. 40,43,46-51 However, one pediatric study calculated a 6.8% rate of significant
377

adiograph misinterpretation. 53 The most common error is failure to recognize an extremity
fracture. 43,46,51,54 Lufkin and colleagues examined the readings of 16,410 consecutive radiographs
and the effect of emergency physicians’ confidence on the accuracy of their interpretations. 55
When ED physicians were confident in their interpretation, the rate of discordant readings was
1.2%. More importantly, the rate of clinically significant errors was only 0.1%.
Rates of misinterpretation are even higher for CT scans. In one study, ED residents and
attending physicians overlooked new infarcts, mass lesions, and cerebral edema, as well as
parenchymal, subarachnoid, and subdural hemorrhages. 52 The rate of clinically important errors
was 20-25%, with failure to recognize a cerebral infarction the most common. 52 It is unclear how
many patients had adverse outcomes as a result, but the authors estimated that less than 1% of
patients were managed inappropriately by ED staff. 52 Other studies confirm that the radiographic
accuracy of non-radiologists in detecting a major cerebral infarction is poor, with
misinterpretation rates of 30-40%. 56-59 This may directly impact patient outcomes since
management of suspected stroke, specifically determining eligibility for thrombolytic therapy,
requires an immediate and accurate interpretation of the CT. In addition, intraobserver and
interobserver reliability are fair to poor, with kappa (κ) values ranging from 0.41-0.20. 57,59
Radiology residents have better accuracy in interpretation of cranial CT scans, though not
as high as that of certified radiologists. As part of a departmental quality control program, Lal
and colleagues found the rate of significant misinterpretations by radiology residents was 0.9%. 60
Roszler et al reported moderate or major errors in 2.1% of resident interpretations of posttraumatic
head CT scans. 61 Wysoki and colleagues 62 found that major discrepancies between the
interpretations of residents and staff radiologists occurred in 1.7% of neuroradiologic studies and
that the rate was significantly higher when CT scans were abnormal rather than normal (12.2%
vs. 1.5%). Residents missed 9% of intracranial hemorrhages and 17% of cranial or facial
fractures.
Opportunities for Impact
The accepted standard of care, supported by the ACR, calls for routine review of all
radiologic studies by a radiologist or other qualified physician in a timely fashion. 8 Given the
available data on staffing patterns, 9 we estimate that on-site radiologists are available to interpret
radiographs about half the time. Radiologists generally read any remaining studies the following
day. In academic centers, the percentage of films initially interpreted by residents may range
from 20% to 100%, and the availability of on-call residents is highly variable. 42 It is also unclear
what proportion of academic and community hospitals provides radiographic training courses for
non-radiologists.
Study Designs
Few studies specifically focus on methods to reduce clinically significant
misinterpretations of radiographs and CT scans. No randomized trials have evaluated the
effectiveness of the 3 patient safety practices identified above. Table 35.1 shows 4 prospective
before-after studies (Level 2 design) of educational interventions and quality improvement
initiatives. 38-41 Evidence for the other 2 strategies (initial interpretations by radiology residents
and review of all studies by radiologists) is limited to descriptive studies reporting rates of
discordant interpretations for various groups of physicians (see above).
378

Study Outcomes
Most studies reported the rate of clinically significant misinterpretations (Level 2
outcome). 38,40,41,53,55,63 However, the definition of “clinically significant” was subjective and
varied among reports. One trial reported only the percentage of correct CT interpretations (Level
3), without describing the significance of specific errors. 39
Evidence for Effectiveness of the Practices
Levitt and colleagues 38 found a significant improvement in cranial CT scan
interpretations by ED physicians after a one-hour training course. However, there were several
limitations of this study. The intervention was not tested before implementation, raising concerns
regarding its reliability and reproducibility. Previously published research at the same institution
showed a 38.7% misinterpretation rate of cranial CT scans by ED physicians. 39 Because
physicians were aware of these results, they may have engaged in their own efforts that led to
improvement in CT scan interpretation. The improvement might also be explained by significant
changes in case-mix, which was not measured. A multicenter study by Perron et al 39 showed that
a reproducible educational course significantly improved residents’ ability to interpret CT scans.
However, the study is limited by possible selection bias (participation in the course was
voluntary; post-test data were obtained in only 61 of 83 subjects). Although subjects in the study
by Perron were retested after a 3-month washout period, neither course has been shown to result
in sustained improvement. Quality improvement programs such as those described by Espinosa
and Preston 40,41 successfully reduced the rate of misinterpreted radiographs and number of
callbacks, respectively. Through ongoing feedback and review of misinterpreted radiographs,
Espinosa and colleagues lowered the rate of important errors from 3% to 0.3%, a relative risk
reduction of 90%. The initiative described by Preston led to a 42.9% relative reduction in
callbacks to the ED, although the absolute magnitude was less impressive (0.3% absolute risk
reduction).
We found no studies that compared the readings of ED physicians with those of
radiology residents in a real-world environment. Only one investigation compared their
diagnostic skill in an experimental situation. 18 Eng and colleagues exposed 4 groups of
physicians (ED attendings, ED residents, radiology attendings, and radiology residents) to a
series of 120 radiographs and calculated their receiver operating characteristic (ROC) curves.
The radiographs were selected for their difficulty, and ED physicians had previously
misinterpreted many of them. The area under the ROC curve was 0.15 higher for radiology
faculty than for ED attendings (95% CI: 0.10-0.20) and 0.07 higher for all faculty than all
residents (95% CI: 0.02-0.12). Compared with ED faculty, radiology residents had an additional
area under the ROC curve of 0.08 (95% CI: 0.02-0.14).
379

Potential for Harm
There is a potential for both radiologists 51 and non-radiologists 64 to overread films (ie,
falsely identify non-existent findings), which may result in unnecessary diagnostic testing or
treatment. Since the vast majority of this literature considers the radiologist’s reading to be the
gold standard, the proportion of discrepancies in interpretation that are due to false-positive
readings by radiologists is unclear.
Costs and Implementation
Only one study reported the costs of false-positive readings by ED physicians, which
averaged $85 per false-positive radiograph. 64 Given the paucity of research, it is not possible to
estimate the costs of implementing universal review of emergency films by an on-site
radiologist. Finally, no studies have measured the costs of misreads to the patient and health care
system (eg, transportation, return visits, repeat or unnecessary studies, unnecessary medications)
which could potentially offset staffing costs. The marginal costs of after-hours staffing alone
may be prohibitive, especially for low-volume sites. No information is available to estimate the
cost of establishing a high-quality teleradiology program, or of educational programs for ED
physicians or trainees.
Comment
The rates of misinterpreted radiographs and CT scans are high in many studies. Of
particular concern is the 20-25% rate of clinically significant errors in reading cranial CT scans,
even among experienced emergency physicians and neurologists. Although plain films are
correctly interpreted more than 90% of the time, even “low” error rates of 1% or less are
important given the sheer number of films.
The relative roles of ED physicians, radiology residents, certified radiologists, and other
physicians will depend on their accuracy and their cost. The literature is notable for the relative
dearth of studies of interventions to reduce radiologic misinterpretation by non-radiologists, even
though dozens of studies have documented the problem. Where radiographs must be interpreted
by non-radiologists, there is limited evidence that brief educational interventions and continuous
quality improvement programs improve diagnostic accuracy. (Chapter 54 reviews general issues
surrounding changing practice behavior through education.) It is possible that radiologists’
routine review of plain radiographs may not be cost-effective when the non-radiologist clinician
has a high level of confidence in the initial interpretation, but this has not been rigorously
established. Coverage by radiology residents may add back-up accuracy to the reading of an ED
attending, but Eng’s study was biased toward a difficult set of radiographs that had previously
been misinterpreted by emergency physicians. 18 The added value of a radiology resident’s
interpretation may be much lower in actual practice.
One avenue of research that could yield effective safety practices is human resource
management. Essentially, staffing could be optimized such that specific tests could be triaged to
the individual with the highest diagnostic accuracy. This kind of intervention would require
assessment of both potential coverage gaps and the skill mix of the available labor force. A
formal protocol might be developed and tested to identify coverage vulnerabilities and develop
realistic coverage options based on measured performance of the available personnel at a
particular site. While the value of this strategy is entirely speculative, it draws on general
management science and could be explored more explicitly in health care.
380

Table 35.1. Educational interventions for non-radiologists*
Study Study Setting Intervention Study Design,
Outcomes
Levitt,
1997 38
Espinosa,
2000 41
Preston,
1998 40
Perron,
1998 39
14 ED physicians at a
level 2 trauma center
in California
ED physicians at an
academic hospital in
New Jersey, 1993-99
ED physicians and
radiologists at a 150bed
community
hospital in Louisiana,
1990-95
83 ED residents at 5
academic centers in
the southeast US,
1997-98
One-hour course on
cranial CT interpretation
Training on plain
radiograph interpretation
and ongoing feedback
from radiologists about
their errors
Continuous quality
improvement initiative
with regular review of
film discrepancies
Two-hour course on
neuroanatomy and cranial
CT interpretation
381
Level 2,
Level 2
Level 2,
Level 2
Level 2,
Level 2
Level 2,
Level 3
Results†
Clinically significant
misinterpretations
decreased from 23.6% to
4.0%
(ARR 19.6%, RRR 83%)
Clinically significant false
negative interpretations:
from 3% to 0.3%
(ARR 2.7%, RRR 90%)
Patient callbacks to ED for
clinically significant
misinterpretation: from
0.7% to 0.4%
(ARR 0.3%, RRR 42.9%)
Correct interpretation of a
series of 12 CT scans:
baseline 60%; 3 months
after the course, 78%
(p

10. Goh KY, Tsang KY, Poon WS. Does teleradiology improve inter-hospital management of
head injury? Can J Neurol Sci.1997;24:235-239.
11. Goh KY, Lam CK, Poon WS. The impact of teleradiology on the inter-hospital transfer of
neurosurgical patients. British Journal of Neurosurgery 1997;11:52-56.
12. Gray WP, Somers J, Buckley TF. Report of a national neurosurgical emergency
teleconsulting system. Neurosurgery 1998;42:103-108.
13. Franken EA, Berbaum KS, Smith WL, Chang PJ, Owen DA, Bergus GR. Teleradiology for
rural hospitals: analysis of a field study. J Telemed Telecare 1995;1:202-208.
14. Franken EA, Berbaum KS, Brandser EA, D'Alessandro MP, Schweiger GD, Smith ML.
Pediatric radiology at a rural hospital: value of teleradiology and subspecialty consultation.
AJR 1997;168:1349-1352.
15. Franken EA, Harkens KL, Berbaum KS. Teleradiology consultation for a rural hospital:
patterns of use. Acad Radiol 1997;4:492-496.
16. Lambrecht CJ. Emergency physicians' roles in a clinical telemedicine network. Ann Emerg
Med 1997;30:670-674.
17. Lee JK, Renner JB, Saunders BF, Stamford PP, Bickford TR, Johnston RE, et al. Effect of
real-time teleradiology on the practice of the emergency department physician in a rural
setting: initial experience. Acad Radiol 1998;5:533-538.
18. Eng J, Mysko WK, Weller GER, et al. Interpretation of emergency department
radiographs: a comparison of emergency medicine physicians with radiologists, residents
with faculty, and film with digital display. AJR 2000;175:1233-1238.
19. Gale ME, Vincent ME, Robbins AH. Teleradiology for remote diagnosis: a prospective
multi-year evaluation. J Digit Imaging 1997;10:47-50.
20. DeCorato DR, Kagetsu NJ, Ablow RC. Off-hours interpretation of radiologic images of
patients admitted to the emergency department: efficacy of teleradiology. AJR
1995;165:1293-1296.
21. Scott WW, Rosenbaum JE, Ackerman SJ, Reichle RL, Magid D, Weller JC, et al. Subtle
orthopedic fractures: teleradiology workstation versus film interpretation. Radiology
1993;187:811-815.
22. Scott WW, Bluemke DA, Mysko WK, Weller GE, Kelen GD, Reichle RK, et al.
Interpretation of emergency department radiographs by radiologists and emergency
medicine physicians: teleradiology workstation versus radiograph readings. Radiology
1995;195:223-229.
23. Calder LD, Maclean JR, Bayliss AP, Gilbert FJ, Grant AM. The diagnostic performance of
a PC-based teleradiology link. Clin Radiol 1999;54:659-664.
24. American College of Radiology. ACR standard for teleradiology; 1998.
25. O'Reilly S, Spedding R, Dearden C, Loane M. Can x-rays be accurately interpreted using a
low-cost telemedicine system? Emerg Med J 1998;15:312-314.
26. Berry RF, Barry MH. Evaluation of a personal-computer-based teleradiology system
serving an isolated Canadian community. Can Assoc Radiol J 1998;49:7-11.
27. Goldberg MA. Teleradiology and telemedicine. Radiol Clin North Am. 1996;34:647-665.
28. Boland GWL. Teleradiology: another revolution in radiology? Clin Radiol 1998;53:547-
553.
29. Baker SR, Festa S. The use of teleradiology in an emergency setting with speculation on
the near future. Radiol Clin North Am 1999;37:1035-1044.
30. Benger J. A review of telemedicine in accident and emergency: the story so far. J Accid
Emerg Med 2000;17:157-164.
382

31. Herman PG, Gerson DE, Hessel SJ, Mayer BS, Watnick M, Blesser B, et al. Disagreements
in chest roentgen interpretation. Chest 1975;68:278-282.
32. Herman PG, Hessel SJ. Accuracy and its relationship to experience in the interpretation of
chest radiographs. Invest Radiol 1975;10:62-67.
33. Hillman BJ, Swensson RG, Hessel SJ, Herson DE, Herman PG. The value of consultation
among radiologists. Am J Roentgenol 1976;127:807-809.
34. Hillman BJ, Hessel SJ, Swensson RG, Herman PG. Improving diagnostic accuracy: a
comparison of interactive and Delphi consultations. Invest Radiol 1977;12:112-115.
35. Hessel SJ, Herman PG, Swensson RG. Improving performance by multiple interpretations
of chest radiographs: effectiveness and cost. Radiology 1978;127:589-594.
36. Rhea JT, Potsaid MS, DeLuca SA. Errors of interpretation as elicited by a quality audit of
an emergency radiology facility. Radiology 1979;132:277-280.
37. Markus JB, Somers S, O'Malley BP, Stevenson GW. Double-contrast barium enema
studies: effect of multiple reading on perception error. Radiology 1990;175:155-6.
38. Levitt MA, Dawkins R, Williams V, Bullock S. Abbreviated educational session improves
cranial computed tomography scan interpretations be emergency physicians. Ann Emerg
Med 1997;30:616-621.
39. Perron AD, Huff JS, Ullrich CG, Heafner MD, Kline JA. A multicenter study to improve
emergency medicine residents' recognition of intracranial emergencies on computed
tomography. Ann Emerg Med 1998;32:554-562.
40. Preston CA, Marr JJ, Amaraneni KK, Suthar BS. Reduction of "call-backs" to the
emergency department due to discrepancies in the plain radiograph interpretation. Am J
Emerg Med 1998;16:160-162.
41. Espinosa JA, Nolan TW. Reducing errors made by emergency physicians in interpreting
radiographs: longitudinal study. BMJ 2000;320:737-740.
42. Townsend RR, Manco-Johnson ML. Night call in U.S. radiology residency programs. Acad
Radiol 1995;2:810-815.
43. Warren JS, Lara K, Connor PD, Cantrell J, Hahn RG. Correlation of emergency department
radiographs: results of a quality assurance review in an urban community hospital setting. J
Am Board Fam Pract 1993;6:255-259.
44. Snow DA. Clinical significance of discrepancies in roentgenographic film interpretation in
an acute walk-in area. J Gen Intern Med 1986;1:295-299.
45. Robinson PJA, Wilson D, Coral A, Murphy A, Verow P. Variation between experienced
observers in the interpretation of accident and emergency radiographs. Br J Radiol
1999;72:323-330.
46. Thomas HG, Mason AC, Smith RM, Fergusson CM. Value of radiograph audit in an
accident service department. Injury 1992;23:47-50.
47. Gratton MC, Salomone JA, Watson WA. Clinically significant radiograph
misinterpretations at an emergency medicine residency program. Ann Emerg Med
1990;19:497-502.
48. Walsh-Kelly CM, Melzer-Lange MD, Hennes HM, Lye P, Hegenbarth M, Sty J, et al.
Clinical impact of radiograph misinterpretation in a pediatric ED and the effect of
physician training level. Am J Emerg Med 1995;13:262-264.
49. Klein EJ, Koenig M, Diekema DS, Winters W. Discordant radiograph interpretation
between emergency physicians and radiologists in a pediatric emergency department.
Pediatr Emerg Care 1999;15:245-248.
383

50. Fleisher G, Ludwig S, McSorley M. Interpretation of pediatric x-ray films by emergency
department pediatricians. Ann Emerg Med 1983;12:153-158.
51. Brunswick JE, Ilkhanipour K, Seaberg DC, McGill L. Radiographic interpretation in the
emergency department. Am J Emerg Med 1996;14:346-348.
52. Alfaro D, Levitt MA, English DK, Williams V, Eisenberg R. Accuracy of interpretation of
cranial computed tomography scans in an emergency medicine residency program. Ann
Emerg Med 1995;25:169-174.
53. Simon HK, Khan NS, Nordenberg DF, Wright JA. Pediatric emergency physician
interpretation of plain radiographs: is routine review by a radiologist necessary and costeffective?
Ann Emerg Med 1996;27:295-298.
54. Freed HA, Shields NN. Most frequently overlooked radiographically apparent fractures in
a teaching hospital emergency department. Ann Emerg Med 1984;31:900-904.
55. Lufkin KC, Smith SW, Matticks CA, Brunette DD. Radiologists' review of radiographs
interpreted confidently by emergency physicians infrequently leads to changes in patient
management. Ann Emerg Med 1998;31:202-207.
56. Wardlaw JM, Dorman PJ, Lewis SC, Sandercock PAG. Can stroke physicians and
neuroradiologists identify signs of early cerebral infarction on CT? J Neurol Neurosurg
Psychiatry 1999;67:651-653.
57. Grotta JC, Chiu D, Lu M, et al. Agreement and variability in the interpretation of early CT
changes in stroke patients qualifying for intravenous rtPA therapy. Stroke 1999;30:1528-
1533.
58. Kalafut MA, Schriger DL, Saver JL, Starkman S. Detection of early CT signs of >1/3
middle cerebral artery infarctions : interrater reliability and sensitivity of CT interpretation
by physicians involved in acute stroke care. Stroke 2000;31:1667-1671.
59. Schriger DL, Kalafut M, Starkman S, Krueger M, Saver JL. Cranial computed tomography
interpretation in acute stroke: physician accuracy in determining eligibility for
thrombolytic therapy. JAMA 1998;279:1293-1297.
60. Lal NR, Murray UM, Eldevik OP, Desmond JS. Clinical consequences of
misinterpretations of neuroradiologic CT scans by on-call radiology residents. AJNR
2000;21:124-129.
61. Roszler MH, McCarroll KA, Rashid T, Donovan KR, Kling GA. Resident interpretation of
emergency computed tomographic scans. Invest Radiol 1991;26:374-376.
62. Wysoki MG, Nassar CJ, Koenigsberg RA, Novelline RA, Faro SH, Faerber EN. Head
trauma: CT scan interpretation by radiology residents versus staff radiologists. Radiology
1998;208:125-128.
63. Seltzer SE, Hessel SJ, Herman PG, Swensson RG, Sheriff CR. Resident film
interpretations and staff review. AJR 1980;137:129-133.
64. Walsh-Kelly CM, Hennes HM, Melzer-Lange MD. False-positive preliminary radiograph
interpretations in a pediatric emergency department: clinical and economic impact. Am J
Emerg Med 1997;15:354-356.
384

Chapter 36. Pneumococcal Vaccination Prior to Hospital Discharge
Scott Flanders, MD
University of California, San Francisco School of Medicine
Background
Streptococcus pneumoniae (pneumococcus) is a leading cause of serious communityacquired
infections, especially pneumonia, the sixth leading cause of death in the United States. 1
It causes over 50,000 cases of bacteremia and at least 500,000 cases of pneumonia annually in
the United States. 1-3 Although pneumococcus is an important pathogen in meningitis, bronchitis,
and otitis media, these disease processes will not be discussed in this report.
The epidemiologic rationale for targeting pneumococcal vaccination among hospitalized
patients derives from research showing that two-thirds of patients hospitalized with serious
pneumococcal infections had been hospitalized at least once in the previous 3-5 years. 4-6 One
retrospective study showed that about 60% of persons 65 years of age and older hospitalized
with pneumonia had been discharged from a hospital at least once in the prior 4 years. 4 A
prospective cohort study of patients aged ≥65 discharged from the hospital showed that the 5year
probability for readmission with pneumonia was over 7%. 4 Hospitalization, therefore, is a
marker for patients at increased risk of developing subsequent pneumococcal infection. Despite
the scope of the problem and the appeals for action from multiple specialty societies and national
health care organizations, the vaccine is underutilized in the inpatient setting. 5-7 This
underutilization has been attributed to uncertain effectiveness of the vaccine and ineffective
methods of vaccine delivery during hospitalization.
Hospital-based vaccination for patients at high risk of contracting pneumococcal
infections is part of the action plan for adult immunizations developed by the Federal Centers for
Disease Control and Prevention (CDC) and the Health Care Financing Administration (HCFA). 6
It is also endorsed by the CDC’s Advisory Committee on Immunization Practices (ACIP) 5, 6 and
the National Vaccine Advisory Committee. In addition, most national guidelines for the
management of patients hospitalized with community-acquired pneumonia recommend
vaccination against pneumococcus at time of discharge. 1,8
Practice Description
Currently, pneumococcal vaccines contain 23 capsular polysaccharide antigens of S.
pneumoniae (23-valent vaccines). Over 88% of the serotypes that cause invasive disease in the
United States, as well as 88% of serotypes accounting for penicillin-resistant isolates, are
included in the 23-valent vaccine. 2,9 Newer conjugate vaccines designed primarily to enhance the
immune response in children are not covered in this review.
The practice of hospital-based pneumococcal vaccination is recommended for patients at
increased risk for pneumococcal infection or increased risk of experiencing severe disease.
Patients ≥65 years of age, or patients with certain chronic illnesses, including chronic
cardiovascular disease, chronic pulmonary disease, diabetes, alcoholism, chronic liver disease,
and functional or anatomic asplenia, are deemed high-risk. 3 It is also recommended that all
immunocompromised patients (due to HIV infection, leukemia, lymphoma, long term steroid
use, or organ transplantation, among other causes) and any patient admitted with a diagnosis of
community-acquired pneumonia be vaccinated. 1,3,8 Vaccination could occur at any time during
the hospitalization, but is often recommended at discharge.
385

Prevalence and Severity of the Target Safety Problem
The goal of pneumococcal vaccination in hospitalized patients is to reduce the morbidity
and mortality associated with pneumococcal infection, namely pneumococcal bacteremia and
pneumococcal pneumonia. The CDC estimates the annual incidence of pneumococcal
bacteremia at 15-30 cases per 100,000 population and 50-83 cases per 100,000 in persons aged
≥65. 3 A recent study of the epidemiology of invasive S. pneumoniae (ie, associated with
bacteremia) in the United States found an overall incidence of 23.2 cases per 100,000,
corresponding to 62,840 cases annually. 2 The incidence among adults aged 65 and older was
59.7 per 100,000. The overall fatality rate was 10%, but patients aged 18-64 with an ACIP
indication for vaccination had a fatality rate of 12.1%. Patients ≥65 years of age accounted for
51.4% of all deaths. These figures result in national estimates of over 6000 deaths in 1998
attributed to invasive pneumococcal disease. 2
The precise incidence of pneumococcal pneumonia is harder to estimate due to the poor
sensitivity and specificity of diagnostic tests for this disease. At least 25-35% of all pneumonias
are linked to S. pneumoniae, resulting in a minimum of 500,000 cases of pneumococcal
pneumonia annually. Bacteremia complicates pneumococcal pneumonia in 10-25% of cases. 3
The mortality rate for all patients hospitalized with community-acquired pneumonia is estimated
at between 10-15%. 1
Opportunities for Impact
Despite recommendations to routinely vaccinate eligible hospitalized patients,
pneumococcal vaccine is underutilized. The high potential impact of vaccination is borne out by
recent epidemiologic evidence. Based on 1998 projections, 76% of invasive pneumococcal
disease and 87% of deaths occur in patients who are eligible for pneumococcal vaccine. 2 In
addition, 88% of penicillin-resistant isolates during the same time period were of serotypes
included in the 23-valent vaccine. 9
The vaccine is currently recommended for over 30 million persons aged ≥ 65 and over 23
million persons

andomized controlled trials (RCTs) of vaccines (valences ranging from 6 to 17) in adults with
and without risk factors for pneumococcal infection. Results were pooled and analyzed for
effects in various subgroups with careful attention to study heterogeneity (rate differences (RD)
were reported when significant heterogeneity existed). 10
The second study 11 included 13 randomized and quasi-randomized studies of vaccines
with valences ≥2. Consequently, this study 11 included 2 quasi-randomized studies with vaccine
valences

overall or in high-risk patients (ie, patients age 55 years, patients with one or more chronic
medical problems, and immunocompromised patients). Results for pneumococcal infectionrelated
outcomes did achieve significance in low-risk patients. 10
Hutchison et al found effectiveness against pneumococcal pneumonia ranging from 31%
to 76% (results were significant in 3 trials), but study heterogeneity prevented a pooled
estimate. 11 Regression analysis suggested similar benefits for systemic pneumococcal infection
in elderly patients, but were inconclusive for systemic infection in chronically ill patients. For
elderly patients, the authors estimated a number needed to treat (NNT) of 2520 to prevent a
single case of pneumococcal bacteremia per year. 11
The most recent meta-analysis 12 reported that in 3 comparisons involving approximately
21,100 immunocompetent subjects, pneumococcal vaccination was associated with significant
reductions in the incidence of all-cause pneumonia (relative risk 0.56, 95% CI: 0.47-0.66),
pneumococcal pneumonia (relative risk 0.16, 95% CI: 0.11-0.23), pneumonia deaths (relative
risk 0.70, 95% CI: 0.50-0.96) and bacteremia (relative risk 0.18, 95% CI: 0.09-0.34). However,
in 10 comparisons involving over 24,000 subjects who were elderly or likely to have impaired
immune systems, the authors found no benefit to pneumococcal vaccination in terms of any
clinical outcome of interest. 12 While the relative risk for pneumococcal bacteremia in elderly or
high-risk patients showed a trend towards benefit, the results were not statistically significant
(relative risk 0.53, 95% CI: 0.14-1.94).
One additional publication 21 has appeared since the search period covered by this most
recent meta-analysis. 12 This publication represents a 6-month preliminary report from a
prospective comparison between 2 large cohorts (>100,000 subjects each) of Swedish patients
age 65 years. Patients in one group received pneumococcal vaccine, influenza vaccine, or both.
The other cohort consisted of all patients from the same region and age group who chose not to
receive either of the vaccines. Among all vaccinated patients (results are pooled for
pneumococcal and influenza vaccines), hospital admission for pneumonia (including all-cause
pneumonia, pneumococcal pneumonia and invasive pneumococcal pneumonia) was significantly
reduced, and overall mortality was reduced by 57% (95% CI: 55-60%). This study design has
significant potential for bias, in that people who elect to participate in clinical studies tend to
have better outcomes independent of the treatments they receive. 22-25 Nonetheless, it seems
unlikely that the results observed in this study, including a 57% reduction in all-cause mortality
over a 6-month period, could be attributable solely to a selection effect tied to patients’ decision
to participate in the study.
36.2. Vaccine Delivery Methods
Study Designs and Outcomes
Multiple studies have evaluated the effectiveness of various methods of increasing rates
of vaccination among eligible patients in both the inpatient and outpatient settings. We focused
our review on interventions relating to inpatient settings (ie, hospitals and nursing homes). Two
systematic reviews published in 1994 and 1999 evaluate multiple strategies. 26, 27 The first review
identified 3 studies in hospitalized patients and one in institutionalized patients. 26 The second
review includes studies identified in the first, and comments on several additional studies. 27 Both
reviews grade the included studies and report pooled estimates of effectiveness. The 2 reviews
use slightly different definitions of the types of interventions, but are internally consistent with
respect to the most effective strategy, namely standing orders.
388

All studies of the effectiveness of vaccine delivery methods reported vaccination rates.
The results of several heterogeneous studies were pooled. There are few RCTs included in the
summary estimates, and most interventions were studied with a before-after analysis of
vaccination rates. Where possible, we report pooled estimates for given methods of improving
vaccine delivery, and then comment on individual studies with the most promise for improving
vaccination rates. No study of delivery methods looked at clinical outcomes.
Evidence for Effectiveness of the Practice
The systematic reviews of vaccine delivery methods with provider reminders in the
inpatient setting were associated with absolute increases in vaccination rates that ranged from
7.5%-17% (Table 36.2.1). 26,27 One subsequent before-after study of chart reminders in
hospitalized patients showed that vaccination rates in eligible patients increased from 0% to
28.8%. 28 The most impressive effects were seen for system-related changes, which increased
vaccination rates by 45-51%. The most effective system-related change was the implementation
of standing orders, which produced increases of 69-81% over usual care. 27 Studies of pneumonia
clinical pathways (Chapter 52) have shown no effect on pneumococcal vaccination rates despite
improvements in other pathway processes. 7
Potential for Harm (from pneumococcal vaccination)
Three of the studies analyzed by Fine et al 10 reported data on adverse effects. Erythema
ranged from 30.6-35.1% in the vaccine group compared with 1.7-3.5% in the control group.
Fever developed in 2.0% of vaccinated patients compared with 1.2% of controls. No fatal or lifethreatening
adverse events occurred. Moore et al 12 report that in one study, in addition to
increased rates of fever, vaccine recipients were more likely to experience a swollen or sore arm.
There is the theoretical concern that patients vaccinated in the hospital may have an increased
chance of being inappropriately revaccinated if they are unaware of their prior vaccination status
(due to illness, etc.) or are being cared for by a physician other than their primary care doctor. A
recent study compared the safety of the vaccine in patients receiving a first vaccination and
patients receiving re-vaccination 5 years after their prior dose. There was an increase in selflimited
local reactions with re-vaccination (RR 3.3, 95% CI: 2.1-5.1), but no serious adverse
reactions were reported. 29 Few data address rates of adverse reactions in hospitalized patients
vaccinated at discharge who receive re-vaccination earlier than 5 years after their prior dose.
Finally, the recent study in Ugandan HIV-infected adults showed trends toward increased
rates of invasive pneumococcal disease, all pneumococcal events, and a statistically significant
increase in all-cause pneumonia (hazard ratio 1.89, 95% CI: 1.1-3.2) among vaccine recipients. 19
This study calls into question the utility of giving pneumococcal vaccine to HIV-infected
individuals.
Costs and Implementation
The cost-effectiveness of hospital-based pneumococcal vaccination is difficult to
determine in light of the debate over how effective the vaccine is in reducing pneumococcal
outcomes in at-risk populations. A cost-effectiveness analysis of vaccination for all elderly
patients in the United States recently demonstrated a wide range of possible outcomes that
depended on assumptions of vaccine effectiveness and duration of protection. 30 Base-case
estimates showed that vaccination was cost saving at $8.27, and gained 1.21 quality-adjusted
days of life per person vaccinated. Factoring in the future medical costs of survivors, vaccinating
all patients ≥ 65 years of age would cost $9600 per quality-adjusted life year under the most
389

optimistic assumptions about vaccine effectiveness, and $51,661 per quality-adjusted life-year
under worst-case assumptions. 30 A recent systematic review of pneumococcal vaccine costeffectiveness
by authors from the Cochrane Vaccines Field concluded that there is too much
variability in assessments of cost-effectiveness (largely attributed to uncertainty over vaccine
effectiveness) to reach any firm conclusions. 31 The authors called for a moratorium on all
economic modeling until completion of a Cochrane review of pneumococcal vaccine
effectiveness.
Comment
Pneumococcal vaccine is effective in reducing invasive disease in low-risk patients. It
appears effective in reducing invasive disease in the elderly and high-risk patients based on
results of one meta-analysis, multiple case-control studies, and a CDC serotype prevalence
study. Importantly however, 2 meta-analyses failed to demonstrate a significant benefit of the
vaccine for any outcomes in elderly or other high-risk patients. The vaccine appears efficacious
in non-bacteremic disease (pneumococcal pneumonia) in low-risk patients, and one metaanalysis
suggests a protective effect against pneumococcal pneumonia in the elderly, 11 but the
others do not. 10,12 No study has demonstrated reductions in mortality. Thus, in the population
most likely to be targeted by hospital-based immunization programs (elderly, high-risk) vaccine
efficacy remains inconclusive. If it is assumed to be effective in reducing the incidence of the
most serious outcome in this population (pneumococcal bacteremia), best estimates suggest a
very large NNT (>2500).
Increasing antibiotic resistance, the aging of the US population and the major burden of
pneumococcal disease among adults and elderly patients make increasing vaccination rates an
obvious goal if, in fact, the vaccine is effective. The evidence supports system changes (in
particular, the use of standing orders) as the best method of increasing vaccine rates in eligible,
hospitalized patients. Though such rates can be increased, available data regarding vaccine
efficacy raise doubts about the overall utility of both local and national initiatives aimed at
increasing the rates of pneumococcal vaccination at hospital discharge. Early enthusiasm
explains the large number of national pneumococcal vaccine initiatives, however available data
on effectiveness provide only modest support for these initiatives.
390

Table 36.1. 1. Pneumococcal vaccine efficacy*
Study Description Study Design,
Outcomes
Meta-analysis of 9 RCTs.
Vaccine valences 6-17,
included international
studies, high- and lowrisk
patients. 10
Meta-analysis of 13 RCTs
and “quasi-randomized”
trials. Vaccine valences 2-
17, includes international
trials, high and low-risk
patients. 11
Meta-analysis of 13 RCTs
including three recent
trials published after the
last meta-analysis. 12
Six-month preliminary
results report from
prospective comparison
between cohort of
100,242 Swedish patients
age 65 years who
received pneumovax,
influenza vaccine, or
both, and patients from
the same region and age
group who chose not to
participate in the study. 21
Level 1A,
Level 1
Level 1A,
Level 1
Level 1A,
Level 1
Level 2,
Level 1
Results (95% Confidence Interval)
* OR indicates odds ratio; RD, risk difference; and RR, relative risk.
Definitive pneumococcal pneumonia:
OR 0.34 (0.24-0.48), RD 4/1000 (0-7)
Definitive pneumococcal pneumonia, vaccine types:
OR 0.17 (0.09-0.33), RD 8/1000 (1-16)
Presumptive pneumococcal pneumonia:
OR 0.47 (0.35-0.63), RD 13/1000 (-21 to 47)
All cause pneumonia: OR 0.90 (0.77-1.04)
Mortality: OR 1.02 (0.90-1.14)
Stratified results show no benefit in any outcome for highrisk
patients.
Systemic infection, vaccine type: OR 0.17 (0.09-0.31)
Systemic infection all types: OR 0.27 (0.13-0.49)
All cause pneumococcal pneumonia: OR range 0.24-0.69,
results significant in 3 studies
Pneumococcal pneumonia, vaccine types:
OR range 0.08-0.85, 8 of 9 studies showed reduced risk,
results significant in 6 studies.
Stratified results show benefit in elderly, mixed results in
chronically ill.
All pneumonias:
Healthy (H), RR 0.56 (0.47-0.66), NNT 29 (24-36);
Elderly, High Risk (E) RR 1.08 (0.92-1.27)
Pneumococcal pneumonias:
H, RR 0.16 (0.11-0.23), NNT 38 (33-45);
E, RR 0.88 (0.72-1.07)
Pneumococcal bacteremia:
H, RR 0.18 (0.09-0.34), NNT 32 (26-44);
E, (3 trials and only 927 pts), RR 0.53 (0.14-1.94)
Pneumonia-related death:
H, RR 0.70 (0.50-0.96), NNT 213 (114-1660);
E, RR 0.93 (0.72-1.20)
Results pool all vaccinated patients.
Hospital admission for all cause pneumonia reduced by 29%
(24-34), for pneumococcal pneumonia 36% (3-58),
invasive pneumococcal pneumonia 52% (1-77).
Overall mortaliy reduced by 57% (55-60)
391

Table 36.2. 1. Vaccine delivery
Study Description Study Design,
Outcomes
Systematic review of studies Level 2A-3A
from 1980-1997 of methods to Vaccination rates
increase vaccination rates. (Level 2)
Multiple vaccines (eg,
pneumococcal, influenza,
hepatitis) delivered in
multiple inpatient and
ambulatory settings were
reviewed, but this summary
focuses on pneumococcal
vaccine in hospitalized
patients
Systematic review of studies Level 2A-3A
from 1979-1992. Multiple Vaccination rates
vaccine types in multiple (Level 2)
settings.
References
392
Results
Provider reminders increased vaccination rates
by 17% (pooled absolute increases for all
vaccines in all settings, with a range of 1-
67%).
Standing orders achieved a 51% mean absolute
increase in vaccination rates (range 30-81%)
for all vaccine types; pneumococcal
vaccination in particular the increases ranged
from 69% to 81% (in hospital setting and longterm
care facility).
Provider-oriented interventions resulted in a
7.5% increase in vaccination coverage (3.4-
11.6%)
System-oriented interventions resulted in a
45.5% increase in vaccination coverage (95%
CI: 37.2-53.7%).
1. Bartlett JG, Dowell SF, Mandell LA, File Jr TM, Musher DM, Fine MJ. Practice guidelines
for the management of community-acquired pneumonia in adults. Infectious Diseases
Society of America. Clin Infect Dis. 2000;31:347-382.
2. Robinson KA, Baughman W, Rothrock G, Barrett NL, Pass M, Lexau C, et al.
Epidemiology of invasive Streptococcus pneumoniae infections in the United States, 1995-
1998: Opportunities for prevention in the conjugate vaccine era. JAMA. 2001;285:1729-
1735.
3. Prevention of pneumococcal disease: recommendations of the Advisory Committee on
Immunization Practices (ACIP). MMWR Morb Mortal Wkly Rep. 1997;46:1-24.
4. Fedson DS, Harward MP, Reid RA, Kaiser DL. Hospital-based pneumococcal
immunization. Epidemiologic rationale from the Shenandoah study. JAMA.
1990;264:1117-1122.
5. Missed opportunities for pneumococcal and influenza vaccination of Medicare pneumonia
inpatients–12 western states, 1995. MMWR Morb Mortal Wkly Rep. 1997;46:919-923.
6. Fedson DS, Houck P, Bratzler D. Hospital-based influenza and pneumococcal vaccination:
Sutton's Law applied to prevention. Infect Control Hosp Epidemiol. 2000;21:692-699.
7. Metersky ML, Fine JM, Tu GS, Mathur D, Weingarten S, Petrillo MK, et al. Lack of effect
of a pneumonia clinical pathway on hospital-based pneumococcal vaccination rates. Am J
Med. 2001;110:141-143.
8. Heffelfinger JD, Dowell SF, Jorgensen JH, Klugman KP, Mabry LR, Musher DM, et al.
Management of community-acquired pneumonia in the era of pneumococcal resistance: a
report from the Drug-Resistant Streptococcus pneumoniae Therapeutic Working Group.
Arch Intern Med. 2000;160:1399-1408.

9. Whitney CG, Farley MM, Hadler J, Harrison LH, Lexau C, Reingold A, et al. Increasing
prevalence of multidrug-resistant Streptococcus pneumoniae in the United States. N Engl J
Med. 2000;343:1917-1924.
10. Fine MJ, Smith MA, Carson CA, Meffe F, Sankey SS, Weissfeld LA, et al. Efficacy of
pneumococcal vaccination in adults. A meta-analysis of randomized controlled trials. Arch
Intern Med. 1994;154:2666-2677.
11. Hutchison BG, Oxman AD, Shannon HS, Lloyd S, Altmayer CA, Thomas K. Clinical
effectiveness of pneumococcal vaccine. Meta-analysis. Can Fam Physician. 1999;45:2381-
2393.
12. Moore RA, Wiffen PJ, Lipsky BA. Are the pneumococcal polysaccharide vaccines
effective? Meta-analysis of the prospective trials. BMC Fam Pract. 2000;1:1.
13. Austrian R. Surveillance of pneumoccal infection for field trials of polyvalent
pneumococcal vaccines. Bethesda MD. 1980: National Institute of Health; 1980. NIH
Publication DAB-VDP-12-84. Contract No 1A13257. 1-59.
14. MacLoed CM, Hodges RG, Heidelberger M, Bernhard WG. CM MacLoed, RG Hodges, M
Heidelberger, WG Bernhard: Prevention of pneumococcal pneumonia by immunization
with specific capsular polysaccharides. J Exp Med. 1945;82:445-465.
15. Kaufman P. Pneumonia in old age: active immunization against pneumonia with
pneumococcus polysaccharide, results of a six year study. Arch Intern Med. 1947;79:518-
531.
16. Honkanen PO, Keistinen T, Miettinen L, Herva E, Sankilampi U, Laara E, et al.
Incremental effectiveness of pneumococcal vaccine on simultaneously administered
influenza vaccine in preventing pneumonia and pneumococcal pneumonia among persons
aged 65 years or older. Vaccine. 1999;17:2493-2500.
17. Koivula I, Sten M, Leinonen M, Makela PH. Clinical efficacy of pneumococcal vaccine in
the elderly: a randomized, single-blind population-based trial. Am J Med. 1997;103:281-
290.
18. Ortqvist A, Hedlund J, Burman LA, Elbel E, Hofer M, Leinonen M, et al. Randomised trial
of 23-valent pneumococcal capsular polysaccharide vaccine in prevention of pneumonia in
middle-aged and elderly people. Swedish Pneumococcal Vaccination Study Group. Lancet.
1998;351:399-403.
19. French N, Nakiyingi J, Carpenter LM, Lugada E, Watera C, Moi K, et al. 23-valent
pneumococcal polysaccharide vaccine in HIV-1-infected Ugandan adults: double-blind,
randomised and placebo controlled trial. Lancet. 2000;355:2106-2111.
20. Butler JC, Breiman RF, Campbell JF, Lipman HB, Broome CV, Facklam RR.
Pneumococcal polysaccharide vaccine efficacy. An evaluation of current
recommendations. JAMA. 1993;270:1826-1831.
21. Christenson B, Lundbergh P, Hedlund J, Ortqvist A. Effects of a large-scale intervention
with influenza and 23-valent pneumococcal vaccines in adults aged 65 years or older: a
prospective study. Lancet. 2001;357:1008-1011.
22. Davis S, Wright PW, Schulman SF, Hill LD, Pinkham RD, Johnson LP, et al. Participants
in prospective, randomized clinical trials for resected non-small cell lung cancer have
improved survival compared with nonparticipants in such trials. Cancer. 1985;56:1710-
1718.
23. Pirie PL, Elias WS, Wackman DB, Jacobs DR, Jr., Murray DM, Mittelmark MB, et al.
Characteristics of participants and nonparticipants in a community cardiovascular disease
risk factor screening: the Minnesota Heart Health Program. Am J Prev Med. 1986;2:20-25.
393

24. Jha P, Deboer D, Sykora K, Naylor CD. Characteristics and mortality outcomes of
thrombolysis trial participants and nonparticipants: a population-based comparison. J Am
Coll Cardiol. 1996;27:1335-1342.
25. Kessenich CR, Guyatt GH, Rosen CJ. Health-related quality of life and participation in
osteoporosis clinical trials. Calcif Tissue Int. 1998;62:189-192.
26. Gyorkos TW, Tannenbaum TN, Abrahamowicz M, Bedard L, Carsley J, Franco ED, et al.
Evaluation of the effectiveness of immunization delivery methods. Can J Public Health.
1994;85(Suppl 1):S14-30.
27. Shefer A, Briss P, Rodewald L, Bernier R, Strikas R, Yusuf H, et al. Improving
immunization coverage rates: an evidence-based review of the literature. Epidemiol Rev.
1999;21:96-142.
28. Vondracek TG, Pham TP, Huycke MM. A hospital-based pharmacy intervention program
for pneumococcal vaccination. Arch Intern Med. 1998;158:1543-1547.
29. Jackson LA, Benson P, Sneller VP, Butler JC, Thompson RS, Chen RT, et al. Safety of
revaccination with pneumococcal polysaccharide vaccine. JAMA. 1999;281:243-248.
30. Sisk JE, Moskowitz AJ, Whang W, Lin JD, Fedson DS, McBean AM, et al. Costeffectiveness
of vaccination against pneumococcal bacteremia among elderly people.
JAMA. 1997;278:1333-1339.
31. Hutton J, Iglesias C, Jefferson TO. Assessing the potential cost effectiveness of
pneumococcal vaccines in the US: methodological issues and current evidence. Drugs
Aging. 1999;15:31-36.
394

395

Chapter 37. Pain Management
Erica Brownfield, MD
Emory University School of Medicine
Approximately 23 million people undergo surgery each year in the United States. 1
Despite pharmacologic interventions, at least 40-50% of postoperative patients report inadequate
pain relief. 2 In addition, the practice of withholding analgesics due to fear of masking
symptomatology and delaying diagnosis is still widespread in many emergency rooms and acute
care settings. The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) and
the Agency for Health Care Policy and Research (AHCPR) have established guidelines for the
appropriate assessment and management of pain in general and postoperatively. Yet efforts to
educate clinicians as to appropriate pain management, particularly in emergency departments
(ED) and following surgery, have lagged behind the available evidence.
We have taken the point of view that untreated pain represents a patient safety problem.
This chapter reviews pain management techniques and interventions in 4 domains: use of
analgesics in patients with acute abdominal pain, the use of acute pain services, prophylactic
antiemetics during patient-controlled analgesia therapy, and non-pharmacologic interventions for
postoperative pain.
Subchapter 37.1. Use of Analgesics in the Acute Abdomen
Background
The use of analgesics in patients with acute abdominal pain has traditionally been
condemned. The 1987 edition of Cope’s Early Diagnosis of the Acute Abdomen states “though it
may appear crude, it is really prudent to withhold morphine until a reasonable diagnosis has been
made and a plan of action formulated.” 1 The most recent edition of Cope’s Early Diagnosis of
the Acute Abdomen (1996) begins to question this long-accepted dogma, but still states that
analgesia medication should be given only after a “responsible surgeon” takes a thorough history
and performs a thorough physical examination. 3 As patients with acute abdominal pain are rarely
evaluated by a surgeon within the first few hours of presentation, it seems inappropriate and
inhumane to withhold pain medication if this practice is not supported by evidence.
Practice Description
Prescribing analgesics to patients with acute abdominal pain is infrequently done. When
prescribed, dosages and routes vary, from intramuscular to intravenous morphine in 5 to 10 mg
increments, or 0.1 mg/kg of body weight. Although JCAHO and AHCPR have established
guidelines for the appropriate assessment and management of pain in general and
postoperatively, neither addresses pain management in patients with acute abdomens. Therefore,
the traditional practice of withholding analgesia in this setting has not been seriously challenged.
Prevalence and Severity of the Target Safety Problem
According to the National Center of Health Statistics, there were 100,385,000 total visits
to US emergency departments in 1998. The most frequent principal reason for visits was
stomach or abdominal pain (5.9 million). 4 Despite studies suggesting that the early
administration of pain medication is safe and does not interfere with, and may actually facilitate,
the ability to make a correct diagnosis, recent surveys of emergency room physicians and
396

surgeons indicate that the majority withhold analgesics in patients presenting with an acute
abdomen. Wolfe et al 2 surveyed 443 emergency medicine physicians and found that although
85% believe that the conservative administration of pain medication did not change important
physical findings, 76% choose not to give an opiate analgesic until after the examination by a
surgeon. Graber et al 5 surveyed 131 practicing surgeons in Iowa and found 67% agreed that pain
medications interfere with diagnostic accuracy, and 82% cited their concerns about diagnostic
accuracy when deciding to withhold pain medication.
Opportunities for Impact
Limited data suggest that the number of patients with acute abdominal pain who actually
receive pain relief before surgical evaluation is small. Therefore, by educating providers in
appropriate pain management for these patients, the potential impact in emergency departments
across the country is large.
Study Designs
Five prospective randomized controlled trials (Level 1) were evaluated. Four of the 5
used a double-blind design. In each study, patients were randomly assigned to receive opiate
analgesia or placebo, and evaluated pre- and post-intervention for pain using variations of visual
analog scales (Table 37.1.1).
Study Outcomes
All 5 studies evaluated the effects of analgesia on pain relief (Level 1), and diagnoses and
treatment decisions (Level 2) in patients with acute abdominal pain. Two studies evaluated the
effects of analgesia on physical examination findings and one evaluated the effects of analgesia
on the diagnostic performance of ultrasonography in patients with acute abdominal pain (Level
3).
Evidence for Effectiveness of the Practice
All 5 studies showed that provision of analgesia decreased pain more than it decreased
localization of tenderness. None of the 5 studies indicate that the practice of providing early
analgesia is harmful. Specifically, no study found compromises in diagnosis or treatment of the
acute abdomen after increasing the use of analgesia.
Potential for Harm
The traditional belief that analgesic use in patients with acute abdominal pain may mask
signs and symptoms, delay diagnosis, and lead to increased morbidity and mortality was not
supported in these studies. All 5 studies analyzed diagnostic or management errors that occurred
in each group.
Attard et al 6 found no difference in localization of physical signs, and no difference in the
surgeon’s diagnostic confidence or management decision (to operate or to observe) between the
2 groups (opioids vs. placebo). The decision to operate or to observe was incorrect in 2 patients
in the opioid group (4%) and in 9 patients in the placebo group (18%). The surgeon’s initial
diagnosis one hour after the injection was incorrect in all of these patients. These same 2 patients
in the opioid group were incorrectly diagnosed as having non-specific abdominal pain when first
assessed, but the diagnosis was subsequently changed and both patients had an inflamed
appendix removed within 24 hours of admission. Neither appendix was perforated. There were
no deaths or side effects from the injection for either group.
397

LoVecchio et al 7 documented changes in localization of physical examination findings
and differences in diagnosis between patients receiving opioid analgesia and placebo. The use of
opioids was associated with some change in tenderness and localization in half the patients but
led to no delays in care or eventual morbidity. The emergency department diagnosis differed
from the final discharge diagnosis in 4 of the 49 total patients. One such patient had received
placebo and 3 had received high-dose morphine (no significant difference). There was no delay
in outcome or time to treatment in any patient.
Zoltie 8 demonstrated that 17/134 (12%) of those receiving opioids had altered physical
signs. Of the 50 patients (out of 288 in the total opiate and placebo groups) whose signs changed
during the evaluation, the most common change (n=32) was alteration in bowel sounds. The
remaining 18 had altered sites of tenderness, in most cases a migration of a large region to a
smaller, more precise area. In no case was the diagnosis altered by a change in physical signs. To
the contrary, the correct diagnosis was facilitated in several cases, particularly in the 18 cases
where the site of pain changed.
Vermeulen et al 9 also found that the use of opioids did not change the appropriateness of
the surgeons’ decision making. Among female patients, the decision to operate was appropriate
more often in the opioid group, but the difference between this group and the placebo group was
not statistically significant. In male patients and overall, opiate analgesia did not influence the
appropriateness of the decision. The appropriateness to discharge patients without surgery was
100% in both groups. No patient who had left the hospital after 24 hours of observation without
surgery was readmitted or operated on at another local hospital. The study also assessed the
impact of analgesia on the accuracy of abdominal sonography. For diagnosis of appendicitis,
ultrasound had lower sensitivity (71.1%) and higher specificity (65.2%) in the opioid group than
in the placebo group, 80.6% and 53.8%, respectively.
Similarly, Pace et al 10 found 3 diagnostic or management errors in each group (out of 35
morphine and 36 control patients). The use of opioids did not alter the physicians’ ability to
evaluate accurately and treat patients appropriately.
Costs and Implementation
The costs associated with implementing appropriate analgesic practice for patients with
acute abdominal pain are limited to physician education programs and the cost of the analgesia
and associated monitoring. There were no cost outcomes reported in any of the 5 studies.
Comment
From the available evidence, we conclude that appropriate use of analgesics in patients
with acute abdominal pain effectively decreases pain and does not interfere with diagnosis or
treatment. Recent surveys suggest many physicians believe conservative administration of pain
medication does not interfere with diagnosis and treatment of patients with acute abdominal
pain. Despite this recognition, the gap between understanding and practice remains large, and
abdominal pain is often undertreated.
398

Table 37.1.1. Randomized controlled trials of analgesia in patients with acute abdominal
pain*
Study
Zoltie, 1986 8
Attard,
1992 6
Pace, 1996 10
LoVecchio,
1997 7
Vermeulen,
1999 9
Study Participants;
Intervention
268 adults with acute
abdominal pain
admitted to a hospital in
the UK; sublingual
buprenorphine vs.
placebo
100 selected adults
admitted with clinically
significant abdominal
pain; intramuscular
papaveretum vs.
placebo
75 patients with acute
abdominal pain at a US
military emergency
department; morphine
49 adults with acute
abdominal pain and
peritoneal signs (“acute
abdomen”) admitted to
the emergency
department of a tertiary
care hospital in NY;
intravenous morphine
vs. placebo
340 adults with pain in
the right lower part of
the abdomen at a
university hospital
emergency department;
intravenous morphine
vs. placebo
Outcomes Results†
Level 1 Pain better after 1 hour: 64/134 vs. 59/122 (p=NS). Only
6/32 (19%) patients who received no tablet reported pain
was better after 1 hour
Change in physical signs after 1 hour: 22/134 (16%) vs.
24/122 (20%); when site of tenderness changed, it usually
was resolution of a large region to a smaller, precise area
In no case was the diagnosis altered by a change in
physical signs
Level 1 Pain score: 3.1 vs. 8.3 (p

References
1. Silen W. Cope’s early diagnosis of the acute abdomen, 17 th ed. New York: Oxford
University Press; 1987.
2. Wolfe JM, Lein DY, Lenkoski K, Smithline HA. Analgesic administration to patients with
an acute abdomen: a survey of emergency medicine physicians. Am J Emerg Med
2000;18:250-153.
3. Silen W. Cope's Early Diagnosis of the Acute Abdomen. 19th ed. New York: Oxford
University Press; 1996.
4. Centers for Disease Control and Prevention. NCHS-FASTATS-Emergency Department
Visits. Available at: http://www.cdc.gov/nchs/fastats/ervisits.htm. Accessed April 1, 2001.
5. Graber MA, Ely JW, Clarke S, Kurtz S, Weir R. Informed consent and general surgeons’
attitudes toward the use of pain medication in the acute abdomen. Am J Emerg Med
1999;17:113-116.
6. Attard AR, Corlett MJ, Kidner NJ, Leslie AP, Fraser IA. Safety of early pain relief for
acute abdominal pain. BMJ 1992;305:554-556.
7. LoVecchio F, Oster N, Sturmann K, Nelson LS, Flashner S, Finger R. The use of
analgesics in patients with acute abdominal pain. J Emerg Med 1997;15:775-779.
8. Zoltie N, Cust MP. Analgesia in the acute abdomen. Ann R Coll Surg Engl 1986;68:209-
210.
9. Vermeulen B, Morabia A, Unger PF, Goehring C, Skljarov I, Terrier F. Acute appendicitis:
influence of early pain relief on the accuracy of clinical and US findings in the decision to
operate – a randomized trial. Radiology 1999;210:639-643.
10. Pace S, Burke TF. Intravenous morphine for early pain relief in patients with acute
abdominal pain. Acad Emerg Med 1996;3:1086-1092.
Subchapter 37.2. Acute Pain Services
Background
The concept of an acute pain service (APS) was first reported in 1988. Its genesis was the
recognition of the problems wrought by inadequate postoperative pain management and
appreciation that acute pain may prolong recovery or precipitate complications. Over the past 15
years, in the United States and worldwide, hospitals have created multidisciplinary acute pain
services, with specially trained staff and resources geared toward providing up-to-date
techniques and education. The APS creates a framework in which postoperative pain can be
managed more effectively, hopefully leading to less discomfort and fewer postoperative
complications.
Practice Description
An APS attempts to bridge the gap between physicians, nurses and patients to coordinate
pain management. The roles of the APS are 1) educating patients, 2) educating nurses and
physicians, 3) selecting appropriate analgesic techniques for different situations, 4) preparing
guidelines for different analgesic regimens, 5) helping to manage acute pain problems, and 6)
performing quality control activities.
Most acute pain services in the United States are anesthesiology-based. The
comprehensive pain management teams usually consist of staff anesthesiologists, resident
400

anesthesiologists, specially trained nurses, pharmacists and physiotherapists. Some services are
nurse-based rather than anesthesia-based.
Prevalence and Severity of the Target Safety Problem
Approximately 50% of patients undergoing surgery do not receive adequate pain relief. 1
Failure to appropriately treat pain stems from lack of knowledge and skills on the part of health
care providers and those responsible for health care system management, and insufficient patient
education. The “safety” problem targeted by the practice of implementing an APS is
postoperative pain and morbidity.
Opportunities for Impact
Approximately 34-44% of hospitals in Europe 2,3 and most major institutions in the United
States 4 have organized APSs. In general, few smaller hospitals have an APS. Therefore, there are
many opportunities to institute acute pain services in hospitals.
Study Designs
Six articles were reviewed for this chapter. All are observational studies; one with a
control, 5 without controls. All 6 studies looked at the intervention of an acute pain service in
patients undergoing surgery (Table 37.2.1).
Study Outcomes
Three of the 6 studies assessed postoperative pain (Level 1). Two of the 6 studies
assessed adverse effects and safety (Level 2) and one assessed knowledge and attitudes,
perceived adequacy of patients’ pain relief and the effect on staff workload and relationships.
Evidence for Effectiveness of the Practice
All 3 studies that assessed postoperative pain scores found improvements. Bardiau et al 2
showed that differences in pain score were most pronounced (around 50%) in patients
undergoing vascular, maxillofacial, gynecologic, oral and urologic surgeries. Gould et al 5
showed a reduction in median visual analog scores for pain during relaxation, movement and
deep inspiration. Tighe et al 6 showed a significant improvement in patient perception of pain
relief after introduction of an APS.
Schug and Torrie 7 found no complications resulting in sustained morbidity or mortality
when anesthesiology-based APS provided postoperative pain relief. Potentially severe
complications (without sequelae) occurred in 0.53% of patients. In one study by Tsui et al, 8 1.8%
of patients developed respiratory complications (bradypnea, hypercapnia, oxygen desaturation),
1.2% developed hypotension, and 28.8% and 15.1%, respectively, developed nausea and
vomiting. None suffered long-term sequelae.
Although the postoperative setting is a logical place for acute pain services, they may
also be useful in patients who experience pain as part of a disease process. Although used more
for managing chronic conditions such as cancer and low back pain, acute pain services are also
gaining popularity in treating hospitalized patients with pain due to a medical condition. There
are no rigorous trials of APSs as they are used for medical patients.
Potential for Harm
Fragmentation of care (ie, lack of continuity between the anesthesiologist performing
preoperative evaluation and anesthesiologist providing postoperative pain control, or
401

fragmentation of care among multiple physicians) and decreased attention by the physician-ofrecord
may result in problems from the intervention. However, no studies have examined these
concerns.
Costs and Implementation
Although none of the studies directly examined costs of implementing an acute pain
service, one study estimated that an APS might be cost-effective. 6 Some data suggest that a
nurse-based APS may be more cost-effective than an anesthesiologist-based APS, although there
are no formal analyses of this supposition. 9
Principal obstacles to implementing such acute pain services include financial
constraints, the challenges of educating newly qualified doctors regarding pain management, and
the complexity of published guidelines. 9
Comment
Studies of APSs are mostly observational, measuring postoperative pain, adverse
outcomes and staff knowledge and attitudes regarding its implementation. Although these studies
indicate that acute pain services can improve postoperative pain without endangering patient
safety, no formal recommendation can be made in the absence of high quality, systematic
reviews of the benefits, costs and feasibility of implementing these services.
402

Table 37.2.1. Studies of acute pain services in postoperative pain management*
Study Setting
1304 patients in the pre-APS
inception phase and 671
patients after its implementation
undergoing various surgeries in
a university teaching hospital 2
2035 patients undergoing
various surgical operations at a
university hospital 5
1518 patients undergoing
various surgeries at a district
general hospital 6
2509 patients under APS care at
a tertiary referral teaching
hospital; 1759 received
systemic analgesia, 590
epidural; 160 other techniques 8
3016 patients treated by an APS
for postoperative pain 7
48 staff members (36 nurses, 12
house officers) working in two
surgical units 10
Study Design,
Outcomes
Level 3,
Level 1
Level 3,
Level 1
Level 3,
Level 1
Level 3,
Level 2
Level 3,
Level 2
Level 3,
Level 3
403
Results
* APS indicates acute pain service; CI, confidence interval.
Significant reduction of all pain indicators
after APS inception (p50%) in patients
undergoing vascular, maxillofacial,
gynecologic, urologic and oral surgeries
Reduction in mean pain from 45 (95% CI:
34-53) to 16 (95% CI: 10-20) after APS
Significant reduction of pain (p

References
1. AHCPR. Acute pain management: operative or medical procedures and trauma, part 1. Clin
Pharm 1992;11:309-331.
2. Bardiau FM, Braeckman MM, Seidel L, Albert A, Boogaerts JG. Effectiveness of an acute
pain service inception in a general hospital. J Clin Anesth 1999;11:583-589.
3. Hall PA, Bowden MI. Introducing an acute pain service. Br J Hosp Med 1996;55:15-17.
4. Ready LB. How many acute pain services are there in the US and who is managing patientcontrolled
analgesia? Anesth 1995;82:322.
5. Gould TH, Crosby DL, Harmer M, Lloyd SM, Lunn JN, Rees GAD, Roberts DE, Webster
JA. Policy for controlling pain after surgery: effect of sequential changes in management.
BMJ 1992;305:1187-1193.
6. Tighe SQ, Bie JA, Nelson RA, Skues MA. The acute pain service: effective or expensive
care? Anaesth 1998;53:382-403.
7. Schug SA, Torrie JJ. Safety assessment of postoperative pain management by an acute pain
service. Pain 1993;55:387-391.
8. Tsui SL, Irwin MG, Wong CML, Fung SKY, Hui TWC, Ng, KFJ, Chan WS, O’Reagan
AM. An audit of the safety of an acute pain service. Anaesth 1997;52:1042-1047.
9. Rawal N. Ten years of acute pain services – achievements and challenges. Reg Anesth Pain
Med 1999;24:68-73.
10. McLeod GA, Davies HTO, Colvin JR. Shaping attitudes to postoperative pain relief: the
role of the acute pain team. J Pain Symptom Manage 1995;10:30-34.
Subchapter 37.3. Prophylactic Antiemetics During Patient-controlled Analgesia Therapy
Background
Nausea and vomiting are common side effects of patient-controlled analgesia (PCA) with
opioids. When severe, nausea and vomiting may limit a patient’s tolerance of PCA. Giving
antiemetics prophylactically has been suggested as a way to prevent PCA-associated nausea and
vomiting.
Practice Description
No study has definitively determined the best prophylactic antiemetic in patientcontrolled
analgesia. Prophylactic treatments that have been used include droperidol, 5-HT3
receptor antagonists (ondansetron, tropisetron), clonidine, promethazine, hyoscine, propofol and
metoclopramide. Prophylactic antiemetics have been given both at the induction of anesthesia
and at the end of surgery. A relatively new approach to minimizing nausea and vomiting
associated with PCA involves the addition of an antiemetic to PCA, so that the patient receives
both analgesic and antiemetic with each PCA demand dose.
Prevalence and Severity of the Target Safety Problem
The incidence of postoperative nausea and vomiting has varied widely (from 8 to 92%)
among studies. 1 There is no evidence that the incidence of nausea and vomiting with PCA is any
different than that with intramuscular opioids. In patients receiving PCA, 57-90% report nausea
and 27-40% report vomiting. There appears to be no clear advantage to using one opioid over
any others in PCA in terms of postoperative emesis.
404

Opportunities for Impact
The degree to which prophylactic antiemetics are used in routine practice with PCA is
unknown.
Study Designs
Tramer and Walder 2 conducted a systematic search for randomized trials (MEDLINE,
Embase, Cochrane library, reference lists, hand-searching, no language restriction) that
compared prophylactic antiemetics with placebo or no treatment in patients receiving
postoperative PCA with opioids. Their review identified 14 placebo-controlled trials with
different regimens of droperidol, ondansetron, hyoscine transdermal therapeutic system,
tropisetron, metoclopramide, propofol and promethazine. One PCA delivered tramadol; all
others delivered morphine.
Both relative risk and number needed to treat were calculated. To estimate the frequency
of drug-related adverse effects, the relative risk and the number needed to harm were calculated.
Study Outcomes
The main end point for efficacy was prevention of emetic events (Level 1). The incidence
of emetic events with active treatments (experimental event rate) and with placebo or no
treatment (control event rate) was extracted. Nausea, vomiting, and “any emetic event” (nausea,
vomiting, or nausea and vomiting) were extracted from each trial. Data on drug-related adverse
effects (Level 2) were extracted as well.
Evidence for Effectiveness of the Practice
Without antiemetic drugs, the incidence of nausea averaged 43% (range 22-80%),
vomiting 55% (45-71%), and any emetic event 67% (54-87%). At 24 hours postoperatively, the
cumulative incidence of nausea and vomiting in patients not receiving any antiemetic treatment
added to their PCA-morphine was approximately 50%.
The best-studied antiemetic was droperidol. It was added to morphine-PCA in 6 placebocontrolled
trials involving 642 patients. Droperidol 0.017-0.17 mg/mg of morphine (0.5-11 mg/d
droperidol) was significantly more effective (p=0.04) in preventing nausea than placebo, without
evidence of dose-responsiveness. Compared with placebo, the number needed to treat with
droperidol to prevent nausea was 2.7 (95% CI: 1.8-5.2) and to prevent vomiting was 3.2 (95%
CI: 2.3-4.8).
The second most frequently reported drugs were 5-HT3 receptor antagonists
(ondansetron, tropisetron). Their effect on vomiting was satisfactory, with numbers needed to
treat of approximately 5 compared with placebo. There was no evidence of any antinausea effect.
Promising results were shown with some of the other interventions (clonidine,
promethazine). However, the limited numbers of patients studied did not generate sufficient
evidence to make a recommendation.
405

Potential for Harm
With placebo, the absolute risk of minor adverse events (sedation, drowsiness and
dizziness or anxiety, restlessness and agitation) was 0-20%. In those trials that reported adverse
events with droperidol, doses of the antiemetic ranged from 1.2 mg to cumulative doses of 7.4
mg. There were no obvious differences in the incidence of minor adverse effects compared with
placebo with droperidol doses

Practice Description
Non-pharmacologic interventions typically fall into 3 categories: health care information,
skills teaching and psychosocial support. 2 Health care information includes information on
preparation for surgery, timing of procedures, functions and roles of health care providers, selfcare
responsibilities, and pain/discomfort information. Skills teaching includes coughing,
breathing and bed exercises, hypnosis, cognitive reappraisal and relaxation exercises. Relaxation
strategies include such techniques as the modified Jacobson method, Flaherty and Fitzpatrick
jaw relaxation, relaxation tapes, tapes with structured breathing, muscle relaxation and pleasant
imagery, and cognitive relaxation. Psychosocial support includes identifying and alleviating
concerns, providing reassurance, problem solving with the patient, encouraging questions and
increasing frequency of support.
Prevalence and Severity of the Target Safety Problem
The safety problem targeted by the practice of non-pharmacologic interventions is
postoperative pain. At least 40-50% of postoperative patients report inadequate pain relief,
despite pharmacologic interventions. 3 Postoperative pain can have deleterious psychological and
physiologic consequences that contribute to patient discomfort and longer recovery periods, and
may compromise outcomes. 1 It also consumes greater health care resources.
Opportunities for Impact
Non-pharmacologic interventions are increasingly used in management of postoperative
pain, although the exact proportion of patients receiving such interventions is unknown.
Study Designs
We identified 7 meta-analyses of studies evaluating the effectiveness of nonpharmacologic
interventions on postoperative pain. Two of these meta-analyses evaluated
various non-pharmacologic interventions for the management of acute pain. Devine 2 conducted a
meta-analysis of 191 studies looking at psychoeducational care (including all 3 intervention
categories above) for adult surgical patients. Sindhu 3 conducted a meta-analysis including 49
randomized controlled trials looking at non-pharmacologic nursing interventions (preoperative
education/information, relaxation, music, imagery, biofeedback, multidisciplinary approaches
and others) for the management of acute pain. AHCPR 1 conducted a systematic search of nondrug
intervention studies on postoperative pain management. One hundred forty studies were
included and formal meta-analysis of 3 was performed. The American Society of
Anesthesiologists Task Force on Pain Management, Acute Pain Section 4 performed a systematic
search for acute pain management in the perioperative setting. Two hundred thirty-three articles
were used in the formal meta-analysis; the number of articles regarding education of patients was
not reported. Good 5 performed a systematic search of trials evaluating the effects of relaxation
and music on postoperative pain. Suls and Wan 6 performed a systematic search of trials
evaluating the effects of sensory and procedural information on coping with stressful medical
procedures and pain. Seers and Carroll 7 performed a systematic review of various relaxation
techniques for postoperative pain relief. Table 37.4.1 lists the 7 articles and briefly describes
their salient features.
407

Study Outcomes
All studies reported postoperative pain (Level 1) as a primary outcome measure. Other
outcomes, such as recovery time, psychological distress, and opioid intake were reported in some
cases (Level 1).
Evidence for Effectiveness of the Practice
The extensive literature covered by the meta-analyses suggested beneficial effects of
psychoeducational care, education and instruction of patients, music and relaxation techniques
(Table 39.5.1). The 2 meta-analyses that examined opioid intake failed to show an effect of nonpharmacologic
interventions in reducing postoperative opioid consumption.
Potential for Harm
One study 7 reported no adverse events in any of the trials for any of the treatment or
control groups. Otherwise, adverse consequences of non-pharmacologic interventions on
postoperative pain management were not addressed.
Costs and Implementation
Direct information on costs of non-pharmacologic interventions was not reported in these
7 studies. Devine and Cook 8 found that the beneficial impact on length of stay and medical
complications rendered non-pharmacologic interventions cost-beneficial. Another review 2
hypothesized that costs could potentially be decreased by non-pharmacologic interventions,
since the length of hospital stay was shortened by an average of 1.5 days (11.5%).
The most obvious direct cost of psychoeducational care is the increased staff time to
provide these services. Based on average treatment duration of 42 minutes, a comprehensive
version of the intervention would probably not take more than 1 hour per patient. 8 Less obvious
direct costs might result from staff time to plan the protocol for psychoeducational care and/or to
develop patient education materials, in-service programs or staff meetings to teach or review the
protocol, printing or purchasing patient education materials, transporting patients to group
teaching sessions, and staff time to document the level of care provided.
Comment
Effective treatment of postoperative pain continues to be a challenge. In addition to
analgesia, non-pharmacologic interventions may provide some benefit in reducing postoperative
pain. Clinicians have a wide range of options to consider when developing a comprehensive
version of non-pharmacologic care appropriate for their patients. As such interventions are low
risk and appeal to many patients, they should be explored in practice and further research.
408

Table 37.4.1. Studies of non-pharmacologic interventions for postoperative pain
Study setting; Practice
Examined
Adult surgical patients (# not
stated), 92% in American hospitals
(45% teaching, 43% general
hospitals); health care information,
skills teaching, psychosocial
support 2
Not reported; transcutaneous nerve
stimulation, education/instruction,
relaxation 1
Not reported; education and
participation of patients and
families in pain control 4
Adult pts who have undergone torso
surgery (# not stated); relaxation
techniques, music 5
Not reported; sensory and
procedural information
(explanations of what will be
happening and how patients can
expect to feel) 6
362 patients undergoing fractured
hip repair, removal of malignant
skin lesions, major elective
abdominal surgery, elective
cholecystectomy, abdominal
hysterectomy, femoral angiography;
jaw relaxation, imagery, music,
breathing, relaxation tapes 7
3387 adult patients; education,
relaxation, music, imagery,
biofeedback, multidisciplinary
approach 3
Study
Design,
Outcomes
Level 1-3 A,
Level 1
Level 1-3 A,
Level 1
Level 1-3 A,
Level 1
Level 1-3 A,
Level 1
Level 1-3 A,
Level 1
Level 1-3 A,
Level 1
Level 1-3 A,
Level 1
409
Results*
79-84% of studies found beneficial effects;
Average effect size values were 0.43 for
recovery, 0.38 for pain, 0.36 for psychological
distress;
Length of stay decreased 11.5%
Simple/complex relaxation techniques,
education/ instruction effective in reducing
mild-moderate pain
Education improves pain control, reduces
adverse outcomes
Total pain decreased in 10 of 13 studies;
Sensory pain† reduction was reported in 6 of 12
studies;
Affective pain‡ decreased in 10 of 13 studies;
Unidimensional pain decreased in 4 of 7 studies;
Observed pain decreased in 4 of 4 studies
Combination of procedural and sensory
preparation significantly better than control on
all measures; effect sizes with combination
larger than with either type of information alone
3 of 7 studies showed significant decrease in
pain sensation/pain distress in those who had
relaxation; 1 out of 5 trials showed significant
improvement in psychological outcomes; less
anxiety in relaxation group
Effect sizes ranged from 2.25 to 1.78; strong
heterogeneity
* Effect size is the standardized difference between the control and experimental groups regarding the
measure of interest in each study. Positive values indicate benefit with the intervention.
† Sensory pain refers to ability to discriminate where pain is occurring and respond appropriately.
‡ Affective pain is the sensation of pain as something unpleasant and to be avoided.

References
1. AHCPR Pain Management Guideline Panel. Clinicians’ quick reference guide to
postoperative pain management in adults. J Pain Symptom Manage 1992;7:214-228.
2. Devine EC. Effects of psychoeducational care for adult surgical patients: a meta-analysis of
191 studies. Patient Educ Couns 1992;19:129-142.
3. Sindhu F. Are non-pharmacological nursing interventions for the management of pain
effective? – a meta-analysis. J Adv Nurs. 1996;24:1152-1159.
4. A Report by the American Society of Anesthesiologists Task Force on Pain Management,
Acute Pain Section. Practice guidelines for acute pain management in the perioperative
setting. Anesth 1995;82:1071-1081.
5. Good M. Effects of relaxation and music on postoperative pain: a review. J Adv Nurs.
1996;24:905-914.
6. Suls J, Wan CK. Effects of sensory and procedural information on coping with stressful
medical procedures and pain: a meta-analysis. J Consult Clin Psychol 1989;57:372-33379.
7. Seers K, Carroll D. Relaxation techniques for acute pain management: a systematic review.
J Adv Nurs 1998;27:466-475.
8. Devine EC, Cook TD. Clinical and cost-saving effects of psychoeducational interventions
with surgical patients: a meta-analysis. Res Nurs Health 1986;9:89-105.
410

Section F. Organization, Structure, and Culture
Chapter 38. "Closed" Intensive Care Units and Other Models of Care for Critically Ill
Patients
Chapter 39. Nurse Staffing, Models of Care Delivery, and Interventions
Chapter 40. Promoting a Culture of Safety
411

412

Chapter 38. “Closed” Intensive Care Units and Other Models of Care for
Critically Ill Patients
Jeffrey M. Rothschild, MD, MPH
Harvard Medical School
Background
Patients in the intensive care unit (ICU) require complex care relating to a broad range of
acute illnesses and pre-existing conditions. The innate complexity of the ICU makes
organizational structuring of care an attractive quality measure and a target for performance
improvement strategies. In other words, organizational features relating to medical and nursing
leadership, communication and collaboration among providers, and approaches to problemsolving
1 may capture the quality of ICU care more comprehensively than do practices related to
specific processes of care. 2
Most features of ICU organization do not exert a demonstrable impact on clinical
outcomes such as morbidity and mortality. 3 While hard clinical outcomes may not represent the
most appropriate measure of success for many organizational features, the role of “intensivists”
(specialists in critical care medicine) in managing ICU patients has shown a beneficial impact on
patient outcomes in a number of studies. For this reason, the Leapfrog Group, representing
Fortune 500 corporations and other large health care purchasers, has identified staffing ICUs
with intensivists as one of three recommended hospital safety initiatives for its 2000 purchasing
principles (see also Chapter 55). 4
In this chapter, we review the benefits of full-time intensivists and the impact of “closed
ICUs” (defined below) on patient outcomes. Much of this literature makes no distinction
between improved outcomes in general and decreased harm in particular. However, given the
high mortality 5 and complication rates 6-8 observed in ICUs, it seems reasonable to consider
global interventions such as organizational changes as patient safety practices.
Practice Description
The following practice definitions are synthesized from studies reviewed for this chapter.
For all of these models, the term “intensivist” refers to a physician with primary training in
medicine, surgery, anesthesiology or pediatrics followed by 2-3 years of critical care medicine
(CCM) training.
Open ICU model—An ICU in which patients are admitted under the care of an internist,
family physician, surgeon or other primary attending of record, with intensivists available
providing expertise via elective consultation. Intensivists may play a de facto primary role in the
management of some patients, but only within the discretion of the attending-of-record.
Intensivist Co-management—An open ICU model in which all patients receive
mandatory consultation from an intensivist. The internist, family physician, or surgeon remains a
co-attending-of-record with intensivists collaborating in the management of all ICU patients.
Closed ICU model—An ICU in which patients admitted to the ICU are transferred to the
care of an intensivist assigned to the ICU on a full-time basis. Generally, patients are accepted to
the ICU only after approval/evaluation by the intensivist. For periods typically ranging from one
week to one month at a time, the intensivist’s clinical duties predominantly consist of caring for
patients in the ICU, with no concurrent outpatient responsibilities.
413

Mixed ICU models—In practice, the above models overlap to a considerable extent.
Thus, some studies avoid attempting to characterize ICUs in terms of these models and focus
instead on the level of involvement of intensivists in patient care regardless of the organizational
model. This involvement may consist of daily ICU rounds by an intensivist (thus including
“closed model ICUs” and “intensivist comanagement”), ICU directorship by an intensivist
(possibly including examples of all 3 models above), or simply the presence of a full-time
intensivist in the ICU (also including examples of all 3 models.)
Intensivist models—ICU management may include all of these models. These models are
contrasted with the open ICU model, in which an intensivist generally does not participate in the
direct care of a significant proportion of the ICU patients.
Prevalence and Severity of the Target Safety Problem
ICUs comprise approximately 10% of acute care hospital beds. 9 The number of annual
ICU admissions in the United State is estimated to be 4.4 million patients. 10 Due to an aging
population and the increasing acuity of illness of hospitalized patients, both the total number of
ICU patients and their proportional share of hospital admissions overall are expected to grow. 11
ICU patients have, on average, mortality rates between 12 and 17%. 25 Overall,
approximately 500,000 ICU patients die annually in the United States. A recent review estimated
that this mortality could be reduced by 15 to 60% using an intensivist model of ICU
management. 12
Young and Birkmeyer have provided estimates of the relative reduction in annual ICU
mortalities resulting from conversion of all urban ICUs to an intensivist model of management
model. 10 Using conservative estimates for current ICU mortality rates of 12%, and estimating
that 85% of urban ICUs are not currently intensivist-managed, the authors calculated that
approximately 360,000 patients die annually in urban ICUs without intensivists. A conservative
projection of a 15% relative reduction in mortality resulting from intensivist-managed ICUs
yields a predicted annual saving of nearly 54,000 lives.
By only measuring ICU mortality rates, this analysis may underestimate the importance
of intensivist-managed ICUs. In addition to mortality, other quality of care outcome measures
that might be improved by intensivists include rates of ICU complications, inappropriate ICU
utilization, patient suffering, appropriate end-of-life palliative care, and futile care.
Opportunities for Impact
Currently, a minority of ICUs in the United States utilizes the intensivist model of ICU
management. 13 Intensivists are even less frequently found in non-teaching and rural hospitals.
The potential impact of the intensivist model is far-reaching.
Study Designs
Among 14 studies abstracted for this chapter, 2 were systematic reviews and 12 were
original studies. One systematic review is an abstract that has not yet appeared in journal form
and does not provide cited references. 12 The other systematic review evaluated 8 references, all
of which are included in this chapter. 10 An additional 4 studies absent from the systematic review
are included here. These 4 studies include 2 abstracts that were published after the 1999
systematic review, 14,15 and 2 studies of pediatric ICUs with intensivists. 16,17
Among the original studies, 6 incorporated historical controls and 5 used a crosssectional
approach. One study 18 had both historical and cross-sectional components. The original
studies include 4 studies of adult medical ICUs, 6 studies of adult surgical ICUs and 2 studies of
414

pediatric multidisciplinary ICUs. Intensivist models used by the studies cited for this review
include 4 closed ICUs, 4 mixed ICUs, 3 ICUs with intensivist comanagement and one open ICU.
Several studies were excluded, including abstracts with insufficient data, 19-25 unclear
distinctions in patient management between control groups and intervention (intensivist
managed) groups, 26,27 intensivist models that may have important roles in future practice (eg,
telemedicine consultation with remote management) but are not yet widely available 28,29 and
considerably older studies. 30
Study Outcomes
Required outcomes of interest in studies chosen for this chapter were ICU mortality,
overall in-hospital mortality, or both. Some studies also included morbidity outcomes, adverse
events and resource utilization (eg, length of ICU and hospital stay), levels of patient acuity or
severity of illness (ICU utilization) and levels of high-intensity intervention usage. Studies
addressing the impact of intensivist ICU management on resource utilization without mortality
or outcome data were excluded. There are no data regarding the impact of intensivists.
Evidence for Effectiveness of the Practice
As shown in Table 38.1, most of the studies report a decrease in unadjusted in-hospital
mortality and/or ICU mortality, although this decrease did not reach statistical significance in 3
of the 14 studies. 16,18,31 One study found a statistically insignificant increase in the unadjusted
mortality rates associated with the intensivist model ICU. 32 This study also found that the ratio of
expected-to-actual mortality was reduced in the intensivist-model ICUs. This finding was
associated with a higher severity of illness scores in the intensivist-model ICU population. A
similar finding of significantly improved outcomes after adjusting for severity of illness and
comparing expected-to-actual mortality rates was demonstrated in one pediatric study. 16 Overall,
the relative risk reduction for ICU mortality ranges from 29% to 58%. The relative risk reduction
for overall hospital mortality is 23% to 50%. These results are consistent with those of a
previous systematic review that found a 15% to 65% reduction in mortality rates in intensivistmanaged
ICUs. 10
Data concerning long-term survival (6 and 12 months) for patients cared for in ICUs with
and without intensivist management is not available. Differences in outcomes between closed
ICUs, mixed ICU models and co-managed ICUs are difficult to assess. Studies that have
addressed conversion from an open to a closed model did not utilize full-time intensivists in the
open model study phases. 18,32-34 Therefore it is not clear to what extent improved patient
outcomes resulted only from changes in intensivists’ direct patient care and supervision.
The observational studies evaluating these practices suffer from 2 major limitations. Half
of the studies retrospectively compared post-implementation outcomes with those during an
historical control period. Because none of these studies included a similar comparison for a
control unit that remained open in both time periods, we lack information on secular trends in
ICU outcomes during the time periods evaluated. The other major limitation associated with
comparing mortality rates for ICU patients relates to differences in ICU admission and discharge
criteria under different organizational models. Under the intensivist model, patients are generally
accepted to the ICU only after approval/evaluation by the intensivist. Thus, conversion to an
intensivist model ICU may bring about changes in the ICU patient population that are
incompletely captured by risk-adjustment models and confound comparisons of mortality rates.
Moreover, these changes in ICU admitting practice may exert contradictory effects. For example,
an intensivist model ICU may result in fewer ICU admissions for patients with dismal
415

prognoses, and less futile care for patients already in the ICU. On the other hand, intensivistmanaged
ICUs with stricter admission and discharge criteria may result in a greater overall
acuity of illness for the ICU patients and therefore higher mortality rates.
Potential for Harm
The potential for harm resulting from intensivist management is unclear. Concerns raised
in the literature about intensivist-managed ICUs include the loss of continuity of care by primary
care physicians, insufficient patient-specific knowledge by the intensivist, 35 reduced use of
necessary sub-specialist consultations, and inadequate CCM training of residents who formerly
managed their own ICU patients.
Perhaps more worrisome is the impact that adoption of this practice would have on
physician staffing and workforce requirements. Without a substantial increase in the numbers of
physicians trained in CCM, projected increases in the ICU patient population over the next 30
years will result in a significant shortfall in the intensivist workforce. 11
Costs and Implementation
These studies did not address the incremental costs associated with implementation of
full-time intensivists. Several studies have analyzed resource utilization and length of stay
associated with intensivist-managed ICUs. 13,16,18,19,29,31,32,36 The results of these studies are
variable with respect to costs. Some demonstrate a decrease in ICU expenses. Others found
increased costs, likely due to the increased use of expensive technologies. Still others show little
overall cost differential. The cost-effectiveness and cost-benefit of an intensivist-model ICU
requires further study.
Comment
Outcomes research in critical care is particularly challenging for several reasons. It
typically relies on observational outcomes studies, and must account for the diversity and
complexity of variables measured and controlled for, such as patient-based, disease-based,
provider-based and therapy-based variables. Despite these challenges and limitations, the
literature fairly clearly shows that intensivists favorably impact ICU patient outcomes. What
remains unclear is which intensivist model to recommend—intensivist consultation versus
intensivist co-management versus closed ICUs. Also, we do not know the degree to which the
choice among these models depends on intensivist background – ie, medicine, anesthesiology or
surgery. Finally, because the mechanism of the benefit of intensivist models is unknown, the
degree to which this benefit can be captured by other changes in practice (eg, adoption of certain
evidence-based processes of ICU care) remains unclear.
The major incentive for clarifying these issues concerns the implications for staffing
ICUs in the future. While the evidence supports the beneficial role of full-time intensivists, the
current number of trainees is insufficient to keep pace with the expected increase in the number
of ICU patients. 11 Until we are able to sufficiently increase the size and number of CCM training
programs for physician specialists, complementary solutions for meeting critical care
management demands should be considered. These might include incorporating physicianextenders
such as nurse practitioners and physician assistants with specialized critical care
training, increased participation by hospitalists in care of ICU patients, 37 regionalization of
critical care services, 38 or providing innovative methods to extend intensivists’ expertise to
remote sites through telemedicine consultations. 28 The latter practice seems particularly
promising—a recent time series cohort study found an approximately 33% decrease in severity-
416

adjusted hospital mortality and a nearly 50% decrease in ICU complications when a technologyenabled
remote ICU management program was instituted in a community-based ICU. 28
417

Table 38.1. Intensivist management in the care of critically ill patients*
Study Setting Study
Year
Closed ICU Model
Tertiary care, urban, teaching
hospital; patients with septic
shock; historical control 33
Teaching hospitals (n=2); two
study designs using historical
and concurrent controls 18
Tertiary care, urban, teaching
hospital; historical control 32
Tertiary care, urban, teaching
hospital; historical control 34
Mixed ICU models
ICUs (n=16) with different
characteristics; cross-sectional 16
ICUs (n=39) with different
characteristics; cross-sectional.
Patients with abdominal aortic
surgery 38
ICUs (n=31) with different
characteristics; cross-sectional.
Patients with esophageal
resection 14
ICUs (n=39) with different
characteristics; cross-sectional.
Patients with hepatic resection 15
Community teaching hospital;
historical control 40
Co-managed ICUs
Tertiary care ICU in a teaching
children’s hospital 16
Tertiary care, Canadian teaching
hospital; historical control 39
Tertiary care, urban, teaching
hospital; cross-sectional
comparison (concurrent
control) 31
1982-
1984
1992-
1993
1993-
1994
1995-
1996
1989-
1992
1994-
1996
1994-
1998
1994-
1998
1992-
1994
1983-
1984
1984-
1986
1994-
1995
ICU Type Study
Design,
Outcomes
MICU Level 3,
Level 1
MICU Level 3,
Level 1
MICU Level 3,
Level 1
SICU Level 3,
Level 1
Pediatric
MICU
SICU
Level 3,
Level 1
SICU Level 3,
Level 1
SICU Level 3,
Level 1
SICU Level 3,
Level 1
MICU Level 3,
Level 1
Pediatric
MICU
SICU
Level 3,
Level 3
SICU Level 3,
Level 1
SICU Level 3,
Level 1
418
Intensivist
Intervention
Mortality Relative Risk
Reduction (%)
ICU Hospital
Closed NA 23
Closed NA Retrospective:
19 (p=NS)
Prospective:
26 (p=NS)
Closed NA -38 (p=NS)†
0/E 13‡
Closed 58 50§
Mixed RRR 25
OR 1.5**
NA
Mixed NA OR 3.0§§
Mixed NA RRR 73
OR 3.5**
Mixed NA RRR 81
OR 3.8**
Open 29 28
Co-manage 48 (p=NS) NA
Co-manage 52 31
Co-manage NA 32 (p=NS)
* ICU indicates intensive care unit; MICU, medicalintensive care unit; Mixed, mixed intensivist model (including daily
ICU rounds by an intensivist, the presence of a full-time intensivist, open units with comanagement and closed units
with mandatory consultations or only intensivist management); NA, not available as outcome (was not evaluated); NS,
not stastically significant; and SICU, surgical intensive care unit.
† Negative value indicates an increase in relative risk of mortality.
‡ O/E is observed to expected mortality ratio based risk adjustment
§ Hospital mortality measured 30-days after discharge

RRR is the unadjusted mortality relative risk reduction
** OR is the adjusted odds ratio of increased mortality associated without an intensivist model.
419

References
1. Zimmerman JE, Shortell SM, Rousseau DM, Duffy J, Gillies RR, Knaus WA, et al.
Improving intensive care: observations based on organizational case studies in nine
intensive care units: a prospective, multicenter study. Crit Care Med. 1993;21:1443-1451.
2. Shojania KG, Showstack J, Wachter RM. Assessing hospital quality: a review for
clinicians. Eff Clin Pract. 2001;4:82-90.
3. Mitchell PH, Shortell SM. Adverse outcomes and variations in organization of care
delivery. Med Care. 1997;35(11 Suppl):NS19-NS32.
4. Milstein A, Galvin R, Delbanco S, Salber P, Buck C. Improving the Safety of Health Care:
The Leapfrog Initiative. Eff Clin Pract. 2000;3:313-316.
5. Knaus WA, Wagner DP, Zimmerman JE, Draper EA. Variations in mortality and length of
stay in intensive care units. Ann Intern Med. 1993;118:753-761.
6. Rubins H, Moskowitz M. Complications of Care in a Medical Intensive Care Unit. J Gen
Intern Med. 1990;5:104-109.
7. Giraud T, Dhainaut JF, Vaxelaire JF, Joseph T, Journois D, Bleichner G, et al. Iatrogenic
complications in adult intensive care units: a prospective two-center study. Crit Care Med.
1993;21:40-51.
8. Ferraris VA, Propp ME. Outcome in critical care patients: a multivariate study. Crit Care
Med. 1992;20:967-976.
9. Groeger JS, Guntupalli KK, Strosberg M, Halpern N, Raphaely RC, Cerra F, et al.
Descriptive analysis of critical care units in the United States: patient characteristics and
intensive care unit utilization. Crit Care Med. 1993;21:279-291.
10. Young M, Birkmeyer J. Potential reduction in mortality rates using an intensivist model to
manage intensive care units. Eff Clin Pract. 2000;3:284-289.
11. Angus DC, MB C, Kelley MA, MD, Schmitz RJ, PhD, et al. Current and projected
workforce requirements for care of the critically ill and patients with pulmonary disease:
can we meet the requirements of an aging population? JAMA. 2000;284:2762-7270.
12. Pronovost PJ, Young T, Dorman T, Robinson K, Angus D. Association between ICU
physician staffing and outcomes: A systematic review. Crit.Care Med. 1999;27:A43.
Abstract.
13. Mallick R, Strosberg M, Lambrinos J, Groeger J. The intensive care unit medical director
as manager: impact on performance. Med Care. 1995;33:611-624.
14. Dimick J, Pronovost P, Heitmiller R, Lipsett P. ICU physician staffing is associated with
decreased length of stay, hospital cost, and complications after esophageal resection. Paper
presented at: Surgical Forum, American College of Surgeons, Chicago, Ill; v LI, p 493-495
Oct 22, 2000.
15. Dimick J, Pronovost P, Lipsett P. The effect of ICU physician staffing and hospital volume
on outcomes after hepatic resection. Crit.Care Med. 2000;28:A77 Abstract.
16. Pollack MM, Katz R, Ruttimann UE, Getson PR. Improving the outcome and efficiency of
intensive care: the impact of an intensivist. Crit Care Med. 1988;16:11-17.
17. Pollack MM, MD, Cuerdon TT, PhD, Patel KM, Ruttimann UE, et al. Impact of quality-ofcare
factors on pediatric intensive care unit mortality. JAMA. 1994;272:941-946.
18. Multz AS, Chalfin DB, Samson IM, Dantzker DR, Fein AM, Steinberg HN, et al. A
"closed" medical intensive care unit (MICU) improves resource utilization when compared
with an "open" MICU. Am J Respir Crit Care Med. 1998;157:1468-1473.
420

19. DiCosmo BFM. Addition of an intensivist improves ICU outcomes in a non-teaching
community hospital. Chest. 1999;116:238S.
20. Jacobs M, Hussain E, Hanna A, Ruskin G, Weiss S, Skrzpiec W. Improving the outcome
and efficiency of surgical intensive care:The impact of full time medical intensivists.
Chest. 1998;114:276S-277S.
21. Render ML, Deddens JA, Thomas B, Wexler LF, Rouan GW. Decreased mortality among
patients cared for in a closed intensive care unit setting. J Gen Intern Med. 1998;13 S 1:19.
22. Marini C, Nathan I, Ritter G, Rivera L, Jurkiewicz A, Cohen J. The impact of full-time
surgical intensivists on ICU utilization and mortality. Crit Care Med. 1995;23(Suppl 1):
A235.
23. Cowen J, Matchett S, Kaufman J, Baker K, Wasser T. Progressive reduction in severityadjusted
mortality after implementation of a critical care program. Crit Care Med.
1999;27(Suppl 1):35A.
24. Tang G, Kuo H. Effect of a full time critical care specialist on ICU mortality. Crit Care
Med. 1996;24:A37.
25. Al-Asadi L, Dellinger R, Deutch J, Nathan S. Clinical impact of closed versus open
provider care in a medical intensive care unit. American Journal of Respiratory & Crit
Care Med. 1996;153:A360.
26. Cole L, Bellomo R, Silvester W, Reeves J. A prospective, multicenter study of the
epidemiology,management, and outcome of severe acute renal failure in a "closed" ICU
system. Am J Resp Crit Care Med. 2000;162:191-196.
27. Blunt M, Burchett K. Out-of-hours consultant cover and case-mix-adjusted mortality in
intensive care. Lancet. 2001;356:735-736.
28. Rosenfeld BA, MD FCCM, Dorman TM, Breslow MJ, Pronovost PM, PhD, Jenckes MM,
et al. Intensive care unit telemedicine: Alternate paradigm for providing continuous
intensivist care. Crit Care Med. 2000;28:3925-3931.
29. Carlson RW, Weiland DE, Srivathsan K. Does a full-time, 24-hour intensivist improve care
and efficiency? Crit Care Clin. 1996;12:525-551.
30. Li TC, Phillips MC, Shaw L, Cook EF, Natanson C, Goldman L. On-site physician staffing
in a community hospital intensive care unit. Impact on test and procedure use and on
patient outcome. JAMA. 1984;252:2023-2027.
31. Hanson CW, 3rd, Deutschman CS, Anderson HL, 3rd, Reilly PM, Behringer EC, Schwab
CW, et al. Effects of an organized critical care service on outcomes and resource
utilization: a cohort study. Crit Care Med. 1999;27:270-274.
32. Carson SS, MD, Stocking CP, Podsadecki TM, Christenson JM, Pohlman AM, et al.
Effects of organizational change in the medical intensive care unit of a teaching hospital: a
comparison of 'open' and 'closed' formats. JAMA. 1996;276:322-328.
33. Reynolds H, Haupt M, Thill-Baharozian M, Carlson R. Impact of critical care physician
staffing on patients with septic shock in a university hospital medical intensive care unit.
JAMA. 1988;260:3446-3450.
34. Ghorra SM, Reinert SE, Cioffi WM, Buczko GM, Simms H, Hank MD. Analysis of the
effect of conversion from open to closed surgical intensive care unit. Ann Surg.
1999;229:163-171.
35. Trunkey D. An unacceptable concept. Arch Surg. 1999;229:172-173.
36. Dicosmo B. Strict ICU admission/discharge criteria result in decreased admissions, shorter
stays and lower costs. Chest. 1999;116:238S-239S.
421

37. Wachter, RM. An introduction to the hospitalist model. Ann Intern Med. 1999; 134:338-
342.
38. Thompson DR, Clemmer TP, Applefeld JJ, Crippen DW, Wedel SK. Regionalization of
Critical Care Medicine. Crit Care Med. 1995;23:1154-1155.
38. Pronovost PJ, Jenckes MW, Dorman T, Garrett E, Breslow MJ, Rosenfeld BA, et al.
Organizational characteristics of intensive care units related to outcomes of abdominal
aortic surgery. JAMA. 1999;281:1310-1317.
39. Brown J, Sullivan G. Effect on ICU Mortality of a Full-time Critical Care Specialist. Chest.
1989;96:127-129.
40. Manthous CA, Amoateng-Adjepong Y, al-Kharrat T, Jacob B, Alnuaimat HM, Chatila W,
et al. Effects of a medical intensivist on patient care in a community teaching hospital.
Mayo Clin Proc. 1997;72:391-399.
422

Chapter 39. Nurse Staffing, Models of Care Delivery, and Interventions
Jean Ann Seago, PhD, RN
University of California, San Francisco School of Nursing
Background
Unlike the work of physicians, the work of registered nurses (RNs) in hospitals is rarely
organized around disease-specific populations. Rather, patients are generally grouped by age
and/or intensity of nursing care (eg, pediatrics or intensive care). Adult patients who require the
least amount of nursing care (the largest proportion of hospitalized patients), may be separated
into medical or surgical units but may also be combined on one unit. Because the work of RNs
and other nurses is organized differently than the work of physicians, this chapter explores the
literature related to nursing structure and process variables that may affect outcomes that relate
to patient safety.
Investigations of patient outcomes in relationship to nurses and their professional
responsibilities in hospitals commonly involve structural measures of care 1-4
including numbers
of nurses, number of nurse hours, percentage or ratios of nurses to patients, organization of
nursing care delivery or organizational culture, nurse workload, nurse stress, or qualification of
nurses. Less commonly, studies involve intervention or process measures of care including
studies based on the science of nursing and others using nurses as the intervention. 1-5
The use of
structural variables rather than process measures to study the impact of nursing activities reflects
the greater availability of data relating to the former (often obtainable from administrative
sources) compared with the latter (typically requiring chart review of direct observation). A
number of structural measures have received considerable attention, specifically measures of
staffing levels in the face of major cost cutting and other changes in health care over the past 15-
20 years. In 1996, the Institute of Medicine 6
reported that there were insufficient data to draw
conclusions about the relationship between nurse staffing and inpatient outcomes. However later
studies have revisited this issue, allowing us to review the literature relating patient outcomes to
various measures of nurse staffing levels, such as full time equivalents (FTEs), skill mix
(proportion of RN hours to total hours), or RN hours per patient day.
This chapter does not address patient outcomes as they relate to various “patient
classification systems” (PCSs), although the prevalence of the use of such systems deserves
mention. PCSs predict nursing care requirements at the individual patient level in order to
determine unit staffing, project budgets, define an objective measure for costing out nursing
services, and to maintain quality standards. 8
Although PCSs are used for multiple purposes, they
are an inadequate tool for determining unit staffing on a daily or shift basis. 9-11
In addition, there
are numerous patient classification systems 12-14
and most are specific to one hospital or one
nursing unit. The validity and reliability of PCSs are inconsistent and the systems cannot be
compared with each other. 8-10,15-28
Thus, rather than reviewing studies that analyze various PCS
scores to patient outcomes, we review studies addressing the question of whether or not “safe
thresholds” exist for levels of nursing care.
423

Practice Description
The availability of nurses, the organization of nursing care, and the types of nursing
interventions vary by institution. Structuring nurse staffing (eg, availability of nurses,
organizational models of nursing care) and care interventions to meet “safe thresholds” could be
considered a patient safety practice. However, no studies have evaluated thresholds explicitly.
This chapter reviews the precursor evidence from observational studies about the strength of the
relationship between nursing variables and patient outcomes, so that possible safe thresholds
may be inferred. We assess evidence that relates patient outcomes to:
1) specific numbers, proportions, or ratios of nurses to patients (nurse staffing);
Nurse availability variables generally characterize the number of hours nurses
spend with patients. Typically, the time is not measured for each patient, but
rather averages are measured based on the census of nurses to patients at a
particular point in time. There are several common ways of accounting for this
nurse staffing and no standardized way to measure it (Table 39.1).
2) specific organization of nursing care delivery, nursing models of care, or
organizational culture; Organization of nursing care variables (Table 39.2)
may also include various nursing care delivery models, nursing unit or
hospital culture, or governance structures. An issue of governance that has
been studied by Aiken 29
and others 30
includes how much autonomy a nurse
has to make practice decisions, how much control she has over practice
decisions, how much collaboration occurs between physicians and nurse in
the organization, and communication patterns; and
3) specific nursing interventions; Although nursing interventions are frequently
studied in outpatient setting, 31,32-39
perhaps because these venues provide
nurses more flexibility to make independent decisions, 40-42
studies in the
inpatient setting have included measures of education, training, or retraining
of nurses, providing audit data to nurses, and capturing nurse assessment of
patient outcomes.
The varieties of intevention studies require some comment. Education interventions are
popular in nursing research because they involve less risk than interventions that directly involve
patients and are more readily approved by hospitals and physicians. 43-51
Unfortunately, some
investigators have made the assumption (which led to the failure to measure clinical outcomes)
that increasing nursing knowledge or changing a practice, such as handwashing, automatically
improves outcomes. 52,46,48,53
Because a large part of a nurse’s job is assessment, investigators have used various
nursing assessments as interventions, such as fall risk assessment, pressure ulcer risk assessment,
or identification of patients at high risk for malnutrition, 55-60
to reduce adverse events. In
multidisciplinary protocols, the nursing activity is often assessment, rather than a nursing process
or procedure. 49
Other process-oriented interventions that lack sufficiently rigorous data to evaluate here,
include specialty nurses, 61,62-65
and interventions based on nursing science in the realm of nurse
decision making in acute care hospitals (eg, mouth care to reduce mucositis, nonpharmacutical
424

interventions to reduce pain, nausea and vomiting, increase sleep, and improve wound
healing). 31,66-73
Prevalence and Severity of the Target Safety Problem
The target safety problems are patient adverse events such as mortality and morbidity.
The challenge is to create an optimum practice environment so that nurses can ideally reduce
safety problems.
Commonly studied adverse hospital events such as falls (Chapter 26), medication errors
(Part III, Section A), and pressure ulcers (Chapter 27), are often used as outcome indicators for
nursing practice. 83-90
Less commonly studied are issues related to improving basic symptom
management (eg, symptoms related to poor sleep, nutrition, or physical activity, or anxiety, pain,
distress and discomfort caused by symptoms, or distress caused by diagnostic tests). In the last
decade there has been increasing public and legislative pressure to improve hospital
environments and address some of the heretofore ignored issues. 91-93
Opportunities for Impact
Unfortunately, there is no definitive evidence as to specific thresholds for RN or total
nursing staff hours per patient day, or nursing skill mix for various patient populations or nursing
unit types. The lack of empirical evidence has been problematic for politicians, the public and
the nursing community. Because decisions about nurse staffing do not have a scientific basis and
are instead based on economics and anecdotes, nurse executives and managers are frequently at
odds with staff nurses; especially those represented by labor unions, over staffing. Nurse
executives are charged with providing safe patient care at a responsible cost. The need to
constrain budgets by reducing nursing hours is in conflict with the needs of the unions and, some
allege, in conflict with the needs of patients.
Based in part on some limited data, New York and Massachusetts have passed legislation
requiring formulae to be developed that ensure safe patient care. 95,96
New Jersey has regulations
which state that licensed nurses shall provide at least 65% of the direct care hours and requires
an acuity system for patient classification. 97
California Assembly Bill 394 directs the California
Department of Health Services to establish nurse-to-patient staffing ratios for acute care
hospitals by January 1, 2002. Sixteen states other than California have nurse staffing legislation
on the calendar but have not implemented ratios. 94
Staffing and ratios are items for collective bargaining and contract negotiations in some
areas. 98-104
Registering complains about “unsafe staffing” may be the nurses’ only recourse
unless there is a negotiated agreement between the union and the hospital.
Current utilization of practices using nursing interventions to make an impact on adverse
hospital events is most likely limited due to uncertainty about effectiveness of specific
interventions. Resources necessary for conducting systematic studies of nursing care provided in
hospitals and then implementing the practices found to be helpful are scarce. 105-109
Study Designs
Searches of MEDLINE from 1990, CINHAL from 1966, documents published by the
American Nurses Association, and the Cochrane Collaboration Library identified no randomized
clinical trials or non-randomized controlled trials analyzing nurse staffing and adverse events.
The study designs for nurse availability (Table 39.3) and organization of care (Table 39.4) are
425

Level 2 or 3 designs. Mitchell et al 111
references several randomized trials in her review article.
However, the articles mentioned used advanced practice nurses such as clinical nurse specialists,
or home care visits as the intervention. 62,112,113
The study by Jorgensen et al 114
was set in a
hospital but the comparison was between a specialty stroke unit and a regular care unit. The
difference was between the different organization of stroke treatment, not nurse skill mix. The
studies abstracted are observational studies that are case control, cohort, before-after, or health
services research using data from large public databases.
The study designs for nurse interventions (Table 39.5) vary from Level 1 to 3. Five
studies use education of nurses as the intervention, and an additional 3 studies cover
enhancements to education efforts (ie, providing data to nurses about adverse events in their
units).
Study Outcomes
The studies of structural measures reported Level 1 or 2 outcomes, along with various
other outcomes such as length of stay, patient satisfaction or nurse satisfaction. Most of the
studies corrected for potential confounders and most adjusted outcomes based on patient acuity.
The process measure studies vary between Level 2 and 3 outcomes. The studies also often
included Level 4 outcomes, such as nurse knowledge, but these did not meet inclusion criteria.
Most of the studies used adverse events such as falls, nosocomial infection, pain, phlebitis,
medication errors or pressure ulcers as outcomes.
Evidence for Effectiveness of the Practice
Nurse Staffing
Table 39.4 summarizes the findings of studies exploring measures of nurse availability.
When measured at the hospital level, there is mixed evidence that nurse staffing is related to 30day
mortality. 30,83,115-118
There is scarce but positive evidence that leaner nurse staffing is
associated with unplanned hospital readmission and failure to rescue. 117,119-121
There is strong
evidence that leaner nurse staffing is associated with increased length of stay, nosocomial
infection (urinary tract infection, postoperative infection, and pneumonia), and pressure
ulcers. 122-125
Results are conflicting as to whether richer nurse staffing has a positive effect on patient
outcomes. Although 5 30,89,118,120,129
of the 16 studies in Table 39.3 reported no association
between richer nurse staffing and positive patient outcomes, the other 11 that report an
association tend to be more recent, with larger samples and more sophisticated methods for
accounting for confounders. These studies had various types and acuities of patients and, taken
together, provide substantial evidence that richer nurse staffing is associated with better patient
outcomes. Although the optimum range for acute care hospital nursing staffing is most likely
within these ranges, none of the studies specifically identify the ratios or hours of care that
produce the best outcomes for different groups of patients or different nursing units.
Models of Nursing Care Delivery
The 7 studies in Table 39.4 provide mixed evidence about the relationship between
organization of nursing care and patient outcomes. Aiken et al 29
found that hospitals with
“magnet” characteristics have lower mortality in one study, but not in another, 115
and Shortell et
426

al 30
also does not find an association in ICUs. Seago 79
found a reduction in medication errors
after a change to patient-focused care and Grillo-Peck et al 130
found a reduction in falls after a
change to a RN-UAP (unlicensed assistive personnel) partner model was introduced. The 2
review articles 111,131
reported mixed results about whether nursing models, nurse surveillance or
work environment is associated with patient outcomes. Thus, the evidence is insufficient to
direct practice.
Nursing Interventions
Table 39.5 provides details about studies using nurse interventions. The first 3 studies
provide support for the idea that added education of nurses reduces infection and
thrombophlebitis. The subsequent 2 studies, however, found no difference in bloodstream
infection or medication error before and after added education. The overall evidence indicates
that using education as the sole intervention does not always change patient outcomes.
Educational interventions were related to changes in nurse practices and, in some studies, also
related to decreasing adverse events. 44,47,54
However adding another intervention such as providing
feedback data or benchmarking results, was more likely to be associated with improved patient
outcomes, 55-57
including decreased infection rates, pressure ulcer rates, and fall rates. 55-57
Potential for Harm
The potential for harm of patients associated with structural interventions such as too few
nurses has been documented. 83-85,124,125
Studies involving process interventions such as using
education of nurses, providing data to nurses, and interventions based on nursing science, seem
to have a low probability of harm, but that is as yet unknown.
Costs and Implementation
Few of the abstracted studies mentioned cost, although several measured length of stay as
an outcome variable. Pratt et al 63
found no difference in quality of care measures using a 100%
RN skill mix and an 80% RN skill mix in 2 wards in one hospital in the United Kingdom. The
cost was less with the 80% skill mix but the nurses who worked with less experienced staff
reported an increase in workload and increase in stress. California is faced with impending
legislated minimum nurse staffing ratios in the acute care hospitals. Based on early studies, 149
at
least 40% of California hospitals may see a negative financial effect because of the need to
increase staffing. Additionally, based on a number of predictions, 150,151
there is now, and there
will continue to be, a significant shortage of registered nurses in the US. Thus, implementing any
increase in RN staffing may be very difficult.
One investigator who provided data to nurses as the intervention related to urinary
catheter infection reported an estimated cost savings of $403,000. 55
Another investigator who
also provided data to nurses related to nosocomial pressure ulcer rates estimated implementation
costs but not cost saving. 57
The investigator who studied adding an IV team (specialty nurses)
reported a savings of $53,000/saved life and $14,000/bloodstream infection. Using clean rather
than sterile dressings on open postoperative wounds saved $9.59/dressing with no change in rate
of wound healing. Based on these studies, it is likely that some nursing interventions can save
costs.
427

Comment
The studies evaluated in this review include only medical, surgical and ICU nursing
units. Other data from more specialized units, the outpatient setting, and those pertaining to
subsets of patients tend to mirror the findings of the evidence evaluation, and are cited in this
section alongside those abstracted and presented in the evidence tables.
The relationship of hospital environment to patient outcomes is still being debated.
However, evidence using hospital-level data indicates increasing the percentage of RNs in the
skill mix, increasing RN FTEs or hours per patient day or average daily census is associated with
decreased risk-adjusted mortality. 116,131,152,153
Other studies, also aggregating data to the hospital
level, found that increasing RN hours per patient day is associated with decreased nosocomial
infection rates, 121,154
decreased urinary tract infections, thrombosis and pulmonary complications
in surgical patients, 124
decreased pressure ulcers, pneumonia, postoperative infection and urinary
tract infection. 122,125
Hunt 117
found that decreasing ratios were related to increasing readmission
rates but were not related to mortality rates.
The cost of primary data collection has limited the number of studies using data
aggregated to the individual nursing unit. There is some evidence that decreased nurse-to-patient
ratios in the ICU was associated with an increase in blood stream infections associated with
central venous catheter, 126
while an increase in agency nurses was related to other negative
patient outcomes. 156
A study in the NICU setting found understaffing and overcrowding of
patients led to an outbreak of Enterobactor cloacae. 155
In 42 ICUs Shortell et al. found that low
nurse turnover was related to shorter length of stay 30
; in 65 units an increase in nurse
absenteeism was related to an increase in urinary tract infection and other patient infections but
not to other adverse events. 157
Amaravadi et al 158
found that night nurse-to-patient ratio in ICUs
in 9 hospitals for a select group of patients who had undergone esophagectomy was not
associated with mortality but was associated with a 39% increase in length of stay and higher
pneumonia rates, reintubation rates, and septicimia rates. As noted previously, Blegan et al found
that as the percentage of RNs per total staff (skill mix) increased there was a decrease in
medication errors, decubitus ulcers, and patient complaints up to a skill mix of 85-87% RNs. 83,84
In several studies, increasing skill mix was associated with decreasing falls, length of
stay, postoperative complications, nosocomial pneumonia, pressure ulcer rates, urinary tract
infection, and postoperative infection. 122-125,130
Several studies with varying sample sizes have
found skill mix to be unrelated to mortality. 111,118,159,160
Others have found skill mix to be
unrelated to treatment problems, postoperative complications, unexpected death rates, or
unstable condition at discharge 129
and found no relationship between skill mix or nursing hours
per patient day and medication errors, falls, patient injuries, and treatment errors. 161
In an early
study of primary (all RN) and team (skill mix) nursing care delivery models, there was no
relationship between percent of RNs and quality of care as measured by nurse report 162
and in 23
hospitals in the Netherlands, there was no relationship between RN-to-patient ratio and incidence
of falls. 89
Although mixed, the overall evidence seems to indicate that proportion of RN hours per
total hours and richer RN-to-patient ratios likely do not affect 30-day mortality, may be
428

associated with in-hospital mortality, and are probably associated with adverse events such as
postoperative complications, nosocomial infection, medication errors, falls, and decubitus ulcers.
Based on recent work, nurse staffing was examined in “best practices” hospitals. This
included hospitals recognized by the American Nurses Association’s Magnet Hospital program,
those commended by the Joint Commission on Accreditation of Healthcare Organizations
(JCAHO), those listed in USA Today’s Top 100 Hospitals, those listed in US News and World
Report’s set of high-quality hospitals, those noted for having better than expected mortality for
heart attacks and newborn readmission rates by the Pacific Business Group on Health (PBGH),
and those recognized by the Bay Area Consumer Checkbook for high quality. There is
significant variation in nurse staffing among these best practices hospitals. The staffing data for
best practices hospitals do not consistently demonstrate that hospitals rated highly for quality of
patient care have uniformly richer staffing than do other hospitals. 74
Because units within
hospitals vary widely in nurse staffing and outcomes, results from data aggregated to the hospital
level are difficult to interpret.
At present the literature is insufficient to make a reasoned judgment about organization
of the work environment of nurses. Further work is needed in the area of nurse interventions. If
there truly is to be an emphasis on reducing adverse events in hospitals and creating hospital
environments that promote health and healing, resources for research related to nurses and
nursing interventions must be found.
429

Table 39.1. Measures of nurse staffing
Nurse Staffing
Measure
Definition
Nurse to patient ratio Number of patients cared for by one nurse typically specified by
job category (RN, Licensed Vocational or Practical Nurse-LVN or
LPN); this varies by shift and nursing unit; some researchers use
this term to mean nurse hours per inpatient day
Total nursing staff or
hours per patient day
RN or LVN FTEs per
patient day
Nursing skill (or staff)
mix
Table 39.2. Models of nursing care delivery
Nursing Care Delivery
Models
All staff or all hours of care including RN, LVN, aides counted per
patient day (a patient day is the number of days any one patient
stays in the hospital, ie, one patient staying 10 days would be 10
patient days)
RN or LVN full time equivalents per patient day (an FTE is 2080
hours per year and can be composed of multiple part-time or one
full-time individual)
The proportion or percentage of hours of care provided by one
category of caregiver divided by the total hours of care (A 60% RN
skill mix indicates that RNs provide 60% of the total hours of care)
Definition
Patient Focused Care A model popularized in the 1990s that used RNs as care managers
and unlicensed assistive personnel (UAP) in expanded roles such as
drawing blood, performing EKGs, and performing certain
assessment activities
Primary or Total
Nursing Care
Team or Functional
Nursing Care
Magnet Hospital
Environment/Shared
governance
A model that generally uses an all-RN staff to provide all direct
care and allows the RN to care for the same patient throughout the
patient’s stay; UAPs are not used and unlicensed staff do not
provide patient care
A model using the RN as a team leader and LVNs/UAPs to
perform activities such as bathing, feeding, and other duties
common to nurse aides and orderlies; it can also divide the work by
function such as “medication nurse” or “treatment nurse”
Characterized as “good places for nurses to work” and includes a
high degree of RN autonomy, MD-RN collaboration, and RN
control of practice; allows for shared decision making by RNs and
managers
430

Table 39.3 Structural measures: availability of nurses and patient outcomes (First 11
studies showed positive associations; final 5 studies detected no significant effect)
Study Setting Study
Design,
Outcomes
1. Data were collected form
1,205 consecutively admitted
patients in 40 units in 20 acute
care hospitals and on 820 nurses
in the US 115
2. All patients who developed a
central venous catheter
bloodstream infection during an
infection outbreak period
(January 1992 through
September 1993) and randomly
selected controls. Cohort study:
all SICU patients during the
study period (January 1991
through September 1993) 126
3. 39 nursing units in 11
hospitals for 10 quarters of data
between July, 1993 and
December, 1995 in the US 84
Level 3,
Level 1&3
Level 3,
Level 1
Level 3,
Level 1&2
Availability of
Nurses
0.8 mean nurse/
patient day with a
range of 0.5-1.5
nurses/patient day
1.2 patient/nurse
and 20 nursing
hours per patient
day (HPPD)
1.5 patient/nurse
and 16 nursing
HPPD
2 patient/nurse and
12 nursing HPPD
Proportion of
direct care RN
hours; total direct
care hours;
Up to 87.5% RN
skill mix
431
Effect Size (coefficient, mean
differences, OR)
This measure was significantly
associated with 30-day mortality
(OR .46, 95% CI: 0.22-0.98). An
additional nurse per patient day
reduces the odds of dying by onehalf.
There was a significant
relationship between nurse to
patient ratios and nursing hours
and central venous catheter
bloodstream infection in the SICU.
For 1.2 patients/nurse and 20
HPPD the adjusted odds ratio was
3.95 (95% CI: 1.07-14.54), 1.5
patients/nurse and 16 nursing
HPPD, 15.6 (95% CI: 1.15-211.4),
and for 2 patients/nurse and 12
HPPD, 61.5 (95% CI:1.23-3074).
With patient acuity controlled,
direct care RN proportion of hours
was inversely associated with
medication errors (-0.525 p

4. 42 inpatient units in one 880bed
hospital in the US 83
5. Data from hospital cost
disclosure reports and patient
discharge abstracts from acute
care hospitals in California and
New York for fiscal years 1992
and 1994 125
6. Data from hospital cost
disclosure reports, patient
discharge abstracts and Medicare
data from acute care hospitals in
Arizona, California, Florida,
Massachusetts, New York, and
Virginia for 1996 123
7. Data from hospital cost
disclosure reports, patient
discharge abstracts from acute
care hospitals in California,
Massachusetts, and New York
for 1992 and 1994 122
Level 3,
Level 1&2
Level 3,
Level 1&2
Level 3,
Level 1&2
Level 3,
Level 1&2
8.63 mean total
hours of care;
69% RN skill mix;
up to 85% skill
mix
7.56-8.43 mean
total hours of
care/nursing
intensity weight
(NIW); 67.7% to
70.5% RN skill
mix
5.76 mean
licensed hours of
care/ 83.3% RN
skill mix
7.67-8.43 mean
total hours of care;
67.7-70.5% skill
mix
432
With patient acuity controlled,
direct care RN proportion of hours
was inversely associated with
medication errors/doses (-
0.576, p

8. Data from HCFA Medicare
Hospital Mortality Information
1986 and the American Hospital
Association 1986 annual survey
of hospitals 116
9. Data from the American
Hospital Association 1986
annual survey of hospitals and
medical record reviews from
July 1987 to June 1988 in 6 large
PPOs 128
10. Data from the American
Hospital Association Annual
Survey of Hospitals for 1993 and
the Nationwide Inpatient Sample
from the Agency for Health Care
Policy and Research for 1993
(HCUP-3) 124
Level 3,
Level 1
Level 3,
Level 3
Level 3,
Level 1
0.9 mean
RN/ADC (average
daily census); 60%
skill mix
52.2 (Texas)-
67.6% (California)
skill mix
67.8% mean skill
mix
433
Controlling for hospital
characteristics, number of
RNs/ADC was not significantly
related to adjusted 30-day
mortality rate but proportion of
RNs/all nursing staff was
significantly related to adjusted
30-day mortality rate (adjusted
difference between lower and
upper fourth of hospitals -2.5, 95%
CI: -4.0 to -0.9)
Controlling for hospital
characteristics, number of
RNs/ADC was not significantly
related to problem rate but
proportion of RNs/all nursing staff
was significantly related to lower
problem rates (California lower
rates 3.58, upper rates 2.30
p

11. Data were collected form
March 1 to June 7, 1986 and
included 497 patients 127
12. 390 patients admitted within
1 week after stroke onset in 9
acute care hospitals in The
Netherlands. Surviving patients
were interviewed 6 months poststroke
and asked about falls. Fall
and other patient data were
collected from medical records.
Ward characteristics were
provided by senior nurses. There
is complete data on 349
patients 89
13. 17,440 patients across 42
ICUs in the US 30
14. Data were collected from
April, 1994-March, 1995 from
23 trusts (groups of hospitals) in
Scotland 117
15. Data were collected form the
American Hospital Association
Annual Survey of Hospitals in
1989-1991, the observed and
predicted 30-day post-admission
mortality for patients with a
primary diagnosis of COPD
from the HCFA Hospital
Information Reports from 1989-
1991 and the Medicare Case Mix
Index 118
16. Data from staffing and
accounting records of 60
community hospitals across the
US in 1985, hospital and nursing
unit surveys, 1981 case mix
indexes from the Federal
Register, and the Health Area
Resources File 129
Level 3,
Level 2
Level 3,
Level 2
Level 3,
Level 1-3
Level 3,
Level 1
Level 3,
Level 1
Level 3,
Level 3
Adequate staffing The adequately staffed unit had
fewer complications than the
inadequately staffed unit.
0.04 mean
difference in nurse
to patient ratios
Mean .66
patient/nurse with
a range of 0.31-
1.31
Mean RN FTE
was 1.21 per
patient
RN FTE/100
adjusted
admissions
52% RN skill mix;
33% LPN mean
nursing HPPD was
4.93
434
There was no statistical difference
in falls between case and control
groups in number of nurses or
nurse ratios on any shift. Days
(mean difference -0.06, CI: -
0.51 to 0.39); Evening (mean
difference -0.24, 95% CI: -0.97 to
0.50); Nights (mean difference
1.24, 95% CI: 0.28 to 2.20); All
shifts (mean difference 0.04, 95%
CI, -0.33 to 0.40).
Neither nurse to patient ratio nor
caregiver interaction was found to
be significantly associated with
risk-adjusted mortality.
There was no association between
RN FTE per occupied hospital bed
and mortality
There was no association between
RN FTE/100 adjusted admissions
and 30-day post-admission
mortality for patients with a
primary diagnosis of COPD
None of the staffing variables of
interest were associated with
medication errors, patient injuries,
IV administration errors, or
treatment errors.

Table 39.4 Structural variables: nursing organization models and patient outcomes
Study Setting Study
Design,
Outcomes
Data were collected from 39
"magnet" hospitals, which are
hospitals designated as good
places for nurses to work, and
195 nonmagnet matched
hospitals 29
Data were collected form 1,205
consecutively admitted patients
in 40 units in 20 acute care
hospitals and on 820 nurses in
the US 115
17,440 patients across 42 ICUs
in the US 30
Data were collected at 3 points in
time; 6 month before the
intervention, 6 months, and 12
months after the introduction of
the new model and included the
time between October 1996 to
December 1997 79
Data were collected 6 months
before and 6 months after the
introduction of the new model
and included the time between
January-June, 1992 and January-
June, 1993 130
Level 3,
Level 1
Level 3,
Level 1&2
Level 3,
Level 1-3
Level 3,
Level 2
Level 3,
Level 2
Review article: Pierce, 1997131 Level 3A,
Level 1&2
Organization
of Care/Models
435
Effect Size (coefficient, mean
differences, OR)
Magnet hospitals Magnet hospitals had a 4.6% lower
adjusted Medicare mortality rates
(p=0.026, 95% CI: 0.9-9.4 fewer
deaths per 1,000)
Magnet hospitals
(nurse control
over practice
variable)
Magnet hospitals
(nurse unit
culture captured
in caregiver
interaction
variable)
Patient Focused
Care
RN-UAP
Partnership
similar to Patient
Focused Care
Nursing
Environment
Nurse control over practice was
not significantly associated with
any clinical outcomes, but was
significantly associated with
patient satisfaction (coefficient
0.56 (95% CI: 0.16-97)
Caregiver interaction was not
significantly associated with
clinical outcomes, but was
significantly associated with lower
risk-adjusted length of stay (-0.16,
p

Review article: MEDLINE from
1966-1996, CINAHL from 1982-
1996, Expanded Academic Index
from 1989-1996, search by
author for investigators known to
be working in the field, manual
searches of the bibliographies of
review articles and monographs
(Mitchell) 111
Level 3A,
Level 1&2
Nursing
Environment
436
Mixed results in studies about
whether nursing surveillance,
quality of working environment,
and quality of interaction with
other professionals predict
hospitals with lower mortality.
With more sophisticated risk
adjustment, evidence suggests that
mortality and complications are
related more to patient variables
and adverse events may be more
closely related to organizational
characteristics.

Table 39.5 Process measures: nurse intervention and patient outcomes
Study Setting Study
Design,
Outcomes
Data were collected from 60
hospitalized patients on 1 surgical
service in a university hospital in
Turkey between September 1996
and September 1997 44
2 surgical and 2 medical wards in
one hospital in Sweden were
randomly assigned to either a
control or experimental group. 18
nurses on the experimental wards
and 18 nurse on the control wards;
90 patients on the experimental
wards and 39 patients on the control
wards; 112 Peripheral IVs on the
experimental wards and 60 PIVs on
the control wards 47
One hospital in Spain; all
nosocomial infection data between
March 1982 and December 1990 54
One university hospital in
Washington, DC; all adult patients
with bloodstream Infections
between July 1984 and February
1994 (n=432) 45
One general hospital in Illinois; all
omitted and wrong dose medication
errors between October 1992 and
March 1993 43
Level 2,
Level 2&3
Level 1,
Level 2&3
Level 3,
Level 1
Level 3,
Level 2
Level 3,
Level 2
437
Intervention Effect Size (coefficient, mean
differences, OR)
Added
education to
intervention
group
Added
education to
intervention
group
Added
education to
intervention
group
Added
education
Added
education
Positive colonization of catheter
hub was 68.6% in the control
group and 25% in the intervention
group (chi square=5.75, p

All urinary catheter-patient-days
between January 1995 and
September 1996 in 1 VA hospital 55
Stanford University Hospital; all
pressure ulcers and nosocomial
pressure ulcers during 1992 through
1996 57
8. Stanford University Hospital 52
bed medical surgical unit; all falls
between 1995 through 1996 56
References
Level 3,
Level 2
Level 3,
Level 2
Level 3,
Level 2
438
Provided
infection rate
data to nurses
Provided
nosocomial
pressure rate
data to nurses
plus added
education
Provided fall
rate data to
nurses and
added
education
Pre-intervention there were
32/1000 catheter-patient days
(95% CI: 22.9-43.7); for the 5
quarters post intervention, there
was a significant decrease
(p

10. Finnigan SA, Abel M, Dobler T, Hudon L, Terry B. Automated patient acuity. Linking
nursing systems and quality measurement with patient outcomes. J Nurs Adm. 1993;23:62-
71.
11. Lawson KO, Formella NM, Smeltzer CH, Walters RM. Redefining the purpose of patient
classification [published erratum appears in Nurs Econ 1993 Nov-Dec;11(6):382]. Nurs
Econ. 1993;11:298-302.
12. Cohen MM, O'Brien-Pallas LL, Copplestone C, Wall R, Porter J, Rose DK. Nursing
workload associated with adverse events in the postanesthesia care unit. Anesthesiology.
1999;91:1882-1890.
13. Garfield M, Jeffrey R, Ridley S. An assessment of the staffing level required for a highdependency
unit. Anaesthesia. 2000;55:137-143.
14. Moreno R, Reis Miranda D. Nursing staff in intensive care in Europe: the mismatch
between planning and practice. Chest. 1998;113:752-758.
15. DeGroot HA. Patient classification systems and staffing. Part 2, Practice and process. J
Nurs Adm. 1994;24:17-23.
16. DeGroot HA. Patient Classification System Evaluation: Part 2, System Selection and
Implemention. J of Nurs Admin. 1989;19:24-30.
17. DeGroot HA. Patient classification systems and staffing. Part 1, Problems and promise. J
Nurs Adm. 1994;24:43-51.
18. Davidhizar R, Mallow GE, Bechtel GA, Giger JN. A patient classification system for the
chronic psychiatric patient. Aust N Z J Ment Health Nurs. 1998;7:126-133.
19. Detwiler C, Clark MJ. Acuity classification in the urgent care setting. J Nurs Adm.
1995;25:53-61.
20. Dunbar LJ, Diehl BC. Developing a patient classification system for the pediatric
rehabilitation setting. Rehabil Nurs. 1995;20:328-332.
21. Freund L, Burrows-Hudson S, Preisig P. Development of a patient classification system for
chronic hemodialysis patients. Am J Kidney Dis. 1998;31:818-829.
22. Giovannetti P, Johnson JM. A new generation of patient classification system. JONA.
1990;20:33-40.
23. Godin M. A patient classification system for the hemodialysis setting. Nurs Manage.
1995;26:66-67.
24. Lovett RB, Reardon MB, Gordon BK, McMillan S. Validity and reliability of medical and
surgical oncology patient acuity tools. Oncol Nurs Forum. 1994;21:1709-1717.
25. Lovett RB, Wagner L, McMillan S. Validity and reliability of a pediatric hematology
oncology patient acuity tool. J Pediatr Oncol Nurs. 1991;8:122-130.
26. O'Brien-Pallas L, Irvine D, Peereboom E, Murray M. Measuring nursing workload:
understanding the variability. Nursing Economics. 1997;15:171-182.
27. O'Brien-Pallas LL, Cockerill RW. Satisfaction with nursing workload systems: report of a
survey of Canadian hospitals. Part B. Can J Nurs Adm. 1990;3:23-26.
28. Walts LM, Kapadia AS. Patient classification system: an optimization approach. Health
Care Manage Rev. 1996;21:75-82.
29. Aiken LH, Smith HL, Lake ET. Lower Medicare mortality among a set of hospitals known
for good nursing care. Med Care. 1994;32:771-787.
30. Shortell SM, Zimmerman JE, Rousseau DM, et al. The performance of intensive care units:
does good management make a difference? Medical Care. 1994;32:508-525.
439

31. Dodd MJ, Dibble SL, Miaskowski C, et al. Randomized clinical trial of the effectiveness of
3 commonly used mouthwashes to treat chemotherapy-induced mucositis. Oral Surgery,
Oral Medicine, Oral Pathology, Oral Radiology and Endodontics. 2000;90:39-47.
32. Robertson R, Samuelson C. Should nurse patient ratios be legislated? - Pros and cons -.
Georgia Nursing. 1996;56:2.
33. Shoemake A. Solutions to the home healthcare nursing shortage. Home Healthc Nurse.
1994;12:35-39.
34. Davis MA, Freeman JW. Excess demand and cost relationships among Kentucky nursing
homes. Health Care Financ Rev. 1994;15:137-152.
35. Carr EC. Talking on the telephone with people who have experienced pain in hospital:
clinical audit or research? J Adv Nurs. 1999;29:194-200.
36. Waterman H, Waterman C. Trends in ophthalmology services, nursing skill-mix and
education: 2nd national survey. J Adv Nurs. 1999;30:942-949.
37. Miskella C, Avis M. Care of the dying person in the nursing home: exploring the care
assistants' contribution including commentary by Froggatt K and Sale DNT. European
Journal of Oncology Nursing. 1998;2:80-88.
38. Eliopoulos C. Nurse staffing in long-term care facilities: the case against a high ratio of
RNs. J Nurs Adm. 1983;13:29-31.
39. Gagnon AJ, Edgar L, Kramer MS, Papageorgiou A, Waghorn K, Klein MC. A randomized
trial of a program of early postpartum discharge with nurse visitation. Am J Obstet
Gynecol. 1997;176:205-211.
40. Cumbey DA, Alexander JW. The relationship of job satisfaction with organizational
variables in public health nursing. J Nurs Adm. 1998;28:39-46.
41. Schmidt CE, Gillies DA, Biordi D, Child DA. Marketing the home healthcare agency. Do
nurses and physicians agree? J Nurs Adm. 1990;20:9-17.
42. Jansen PGM, Kerkstra A, Abu Saad HH, van der Zee J. Models of differentiated practice
and specialization in community nursing: a review of the literature. J Adv Nurs.
1996;24:968-980.
43. Carey RG, Teeters JL. CQI case study: reducing medication errors. Jt Comm J Qual
Improv. 1995;21:232-237.
44. Dinc L, Erdil F. The effectiveness of an educational intervention in changing nursing
practice and preventing catheter-related infection for patients receiving total parenteral
nutrition. Int J Nurs Stud. 2000;37:371-379.
45. Cohran J, Larson E, Roach H, Blane C, Pierce P. Effect of intravascular surveillance and
education program on rates of nosocomial bloodstream infections. Heart Lung.
1996;25:161-164.
46. Flynn ER, Wolf ZR, McGoldrick TB, Jablonski RA, Dean LM, McKee EP. Effect of three
teaching methods on a nursing staff's knowledge of medication error risk reduction
strategies. J Nurs Staff Dev. 1996;12:19-26.
47. Lundgren A, Wahren LK. Effect of education on evidence-based care and handling of
peripheral intravenous lines. J Clin Nurs. 1999;8:577-585.
48. Gould D, Chamberlain A. The use of a ward-based educational teaching package to
enhance nurses' compliance with infection control procedures. J Clin Nurs. 1997;6:55-67.
49. Hendryx MS, Fieselmann JF, Bock MJ, Wakefield DS, Helms CM, Bentler SE. Outreach
education to improve quality of rural ICU care. Results of a randomized trial. Am J Respir
Crit Care Med. 1998;158:418-23.
440

50. Kite K. Changing mouth care practice in intensive care: implications of the clinical setting
context. Intensive Crit Care Nurs. 1995;11:203-209.
51. Knoblauch SC, Wilson CJ. Clinical outcomes of educating nurses about pediatric pain
management. Outcomes Manag Nurs Pract. 1999;3:87-89.
52. Spetz J, Coffman J. Maintaining an adequate supply of registered nurses in California: the
case for increased public investment in nursing education [abstract]. Abstract Book /
Association for Health Services Research. 1999:426-427.
53. Barnason S, Merboth M, Pozehl B, Tietjen MJ. Utilizing an outcomes approach to improve
pain management by nurses: a pilot study . Clin Nurse Spec. 1998;12:28-36.
54. Fernandez-Perez C, Tejada J, Carrasco M. Multivariate time series analysis in nosocomial
infection surveillance: a case study. Int J Epidemiol. 1998;27:282-288.
55. Goetz AM, Kedzuf S, Wagener M, Muder RR. Feedback to nursing staff as an intervention
to reduce catheter-associated urinary tract infections. Am J Infect Control. 1999;27:402-
404.
56. Huda A, Wise LC. Evolution of compliance within a fall prevention program. J Nurs Care
Qual. 1998;12:55-63.
57. Moore SM, Wise L. Reducing nosocomial pressure ulcers. J Nurs Adm. 1997;27:28-34.
58. Frykberg RG. The team approach in diabetic foot management. Adv Wound Care.
1998;11:71-77.
59. Gunningberg L, Lindholm C, Carlsson M, Sjoden PO. Implementation of risk assessment
and classification of pressure ulcers as quality indicators for patients with hip fractures. J
Clin Nurs. 1999;8:396-406.
60. Rubenstein LZ, Robbins AS, Josephson KR, Schulman BL, Osterweil D. The value of
assessing falls in an elderly population. A randomized clinical trial. Ann Intern Med.
1990;113:308-316.
61. Meier PA, Fredrickson M, Catney M, Nettleman MD. Impact of a dedicated intravenous
therapy team on nosocomial bloodstream infection rates. Am J Infect Control.
1998;26:388-392.
62. Naylor M, Brooten D, Jones R, Lavizzo-Mourey R, Mezey M, Pauly M. Comprehensive
discharge planning for the hospitalized elderly. A randomized clinical trial. Ann Intern
Med. 1994;120:999-1006.
63. Pratt R, Burr G, Leelarthaepin B, Blizard P, Walsh S. The effects of All-RN and RN-EN
staffing on the quality and cost of patient care. Aust J Adv Nurs. 1993;10:27-39.
64. Ridsdale L, Kwan I, Cryer C. Newly diagnosed epilepsy: can nurse specialists help? A
randomized controlled trial. Epilepsy Care Evaluation Group. Epilepsia. 2000;41:1014-
1019.
65. York R, Brown LP, Samuels P, et al. A randomized trial of early discharge and nurse
specialist transitional follow-up care of high-risk childbearing women. Nursing Research.
1997;46:254-261.
66. Evans ML, Dick MJ, Shields DR, Shook DM, Smith MB. Postpartum sleep in the hospital.
Relationship to taking-in and taking-hold. Clin Nurs Res. 1998;7:379-389.
67. Mock V, Dow KH, Meares CJ, et al. Effects of exercise on fatigue, physical functioning,
and emotional distress during radiation therapy for breast cancer. Oncol Nurs Forum.
1997;24:991-1000.
68. Sheely LC. Sleep disturbances in hospitalized patients with cancer. Oncol Nurs Forum.
1996;23:109-111.
441

69. ter Riet G, Kessels AG, Knipschild PG. Randomized clinical trial of ascorbic acid in the
treatment of pressure ulcers. J Clin Epidemiol. 1995;48:1453-1460.
70. Hardyck C, Petrinovich L. Reducing urinary tract infections in catheterized patients.
Ostomy/Wound Management. 1998;44:36-8, 40, 42-43.
71. Stotts NA, Barbour S, Griggs K, et al. Sterile versus clean technique in postoperative
wound care of patients with open surgical wounds: a pilot study. J Wound Ostomy
Continence Nurs. 1997;24:10-18.
72. Wipke-Tevis DD, Stotts NA. Effect of dressings on saphenous vein harvest incision pain,
distress and cosmetic result. Prog Cardiovasc Nurs. 1998;13:3-13.
73. Gilcreast DM, Stotts NA, Froelicher ES, Baker LL, Moss KM. Effect of electrical
stimulation on foot skin perfusion in persons with or at risk for diabetic foot ulcers. Wound
Repair Regen. 1998;6:434-441.
74. Spetz J, Seago JA, Coffman J, Rosenoff E, O'Neil E. Minimum Nurse Staffing Ratios in
California Acute Care Hosptials. San Francisco, CA: University of California, San
Francisco Center for the Health Professions; 2000.
75. Rogut L, Hudson A. Meeting patients' needs: quality care in a changing environment. Pap
Ser United Hosp Fund N Y. 1995:1-33.
76. Minion M, Ogden C, Brune D. Patient and staff needs drive changes on a postsurgical unit.
Nurs Adm Q. 1994;18:18-28.
77. Managed care and nursing: a view from the front lines. Massachusetts Nurse. 1995;65:4,
10-11.
78. Seago JA, Faucett J. Job strain among registered nurses and other hospital workers. J Nurs
Adm. 1997;27:19-25.
79. Seago JA. Evaluation of work redesign: Patient focused care. J Nurs Adm. 1999;29:31-38.
80. Sochalski J, Aiken LH, Fagin C. Hospital Restructuring in the United States, Canada, and
Western Europe: An Outcomes Research Agenda. Med Care. 1997;35:OS13-25.
81. Sochalski J, Boulis A, Shamian J, Buchan J, Muller-Mundt G. Lessons for restructuring the
nursing workforce in North American and European hospitals [abstract]. Abstract Book /
Association for Health Services Research. 1997;14:207-208.
82. Weinberg DB. Why aare the nurses crying? Restructuring, power, an dcontrol in an
American hospital. Boston, MA: Harvard; 2000.
83. Blegen MA, Goode CJ, Reed L. Nurse staffing and patient outcomes. Nurs Res.
1998;47:43-50.
84. Blegen MA, Vaughn T. A multisite study of nurse staffing and patient occurrences.
Nursing Economics. 1998;16:196-203.
85. McCloskey JM. Nurse staffing and patient outcomes. Nursing Outlook. 1998;46:199-200.
86. Reed L, Blegen MA, Goode CS. Adverse patient occurrences as a measure of nursing care
quality. J Nurs Adm. 1998;28:62-69.
87. Stevenson B, Mills EM, Welin L, Beal KG. Falls risk factors in an acute-care setting: a
retrospective study. Can J Nurs Res. 1998;30:97-111.
88. Sutton JC, Standen PJ, Wallace WA. Patient accidents in hospital: incidence,
documentation and significance. Br J Clin Pract. 1994;48:63-66.
89. Tutuarima JA, de Haan RJ, Limburg M. Number of nursing staff and falls: a case-control
study on falls by stroke patients in acute-care settings. J Adv Nurs. 1993;18:1101-1105.
90. Wolf ZR, McGoldrick TB, Flynn ER, Warwick F. Factors associated with a perceived
harmful outcome from medication errors: a pilot study. J Contin Educ Nurs. 1996;27:65-
74.
442

91. Medicare. Available at: www.state.ia.us/ins/shiip/resources.html. Accessed June 23, 2001.
92. Rhode Island Department of Health. Available at: http://www.health.state.ri.us//. Accessed
June, 23 2001.
93. California Consumer Health Scope. Available at: www.healthscope.org/hospital/where.asp.
Accessed June 30, 2001.
94. Denby C, Hogan N. Retaining Nursing Talent. In: Company TAB, ed. The Journey
Begins. Atlanta, GA; 1999.
95. Ballard K. Legislating Patient Advocacy: Nurse Staffing Ratios. In: Curtin L, ed. A
Staffing Crisis: Nurse/Patient Ratios. Washington, DC; 2000.
96. Ober S. Legislating Patient Advocacy: Nurse Staffing Ratios. In: Curtin L, ed. Staffing
Crisis: Nurse/Patient Ratios. Washington, DC; 2000.
97. 1997 NJSNA annual survey. New Jersey Nurse. 1997;27:7.
98. Staffing is the top prioirty for MNA bargaining units. Massachusetts Nurse. 1999;69:5.
99. MNA members take to the streets for safe patient care. Massachusetts Nurse. 1999;69:12.
100. Staffing is the top priority for MNA bargaining units. Massachusetts Nurse. 1999;69:5.
101. New study confirms critical role of low nurse-to-patient ratio. Massachusetts Nurse.
1999;69:5.
102. Maryland RNs win big in a pact that sets patient-load lilmits. Am J Nurs. 1990;127: 134.
103. Cape Cod Hospital RNs ratify landmark agreement. Massachusetts Nurse. 1997;67: 1-2.
104. Ratio ruckus. Modern Healthcare. 1998;28:48.
105. Alexander JW, Bauerschmidt AD. Implications for nursing administration of the
relationship of technology and structure to quality of care. Nurs Adm Q. 1987;11:1-10.
106. Currell R, Wainwright P, Urquhart C. Nursing record systems: effects on nursing practice
and health care outcomes (Cochrane Review). In: The Cochrane Library, Issue 4, 2000.
Oxford.
107. Curtin LL. Nursing productivity: from data to definition. Nurs Manag . 1995;26:25, 28-29,
32-36.
108. Kovner C, Rosenfeld P. Practice and employment trends among nurse practitioners in New
York State. J N Y State Nurses Assoc. 1997;28:4-8.
109. Park HA, Park JH. Development of a computerized patient classification and staffing
system. Stud Health Technol Inform. 1997;46:508-511.
110. Spetz J. The measurement of technology in studies of the hospital industry [abstract]. Ahsr
and Fhsr Annual Meeting Abstract Book. 1996;13:100.
111. Mitchell PH, Shortell SM. Adverse outcomes and variations in organization of care
delivery. Med Care. 1997;35:NS19-32.
112. Brooten D, Roncoli M, Finkler S, Arnold L, Cohen A, Mennuti M. A randomized trial of
early hospital discharge and home follow-up of women having cesarean birth. Obstet
Gynecol. 1994;84:832-838.
113. Alexander JS, Younger RE, Cohen RM, Crawford LV. Effectiveness of a nurse-managed
program for children with chronic asthma. J Pediatr Nurs. 1988;3:312-317.
114. Jorgensen HS, Kammersgaard LP, Houth J, et al. Who benefits from treatment and
rehabilitation in a stroke Unit? A community-based study. Stroke. 2000;31:434-439.
115. Aiken LH, Sloane DM, Lake ET, Sochalski J, Weber AL. Organization and outcomes of
inpatient AIDS care. Med Care. 1999;37:760-772.
116. Hartz AJ, Krakauer H, Kuhn EM, et al. Hospital characteristics and mortality rates. N Engl
J Med; 1989:1720-1725.
443

117. Hunt J, Hagen S. Occasional paper. Nurse to patient ratios and patient outcomes. Nursing
Times. 1998;94:63-66.
118. Robertson RH, Hassan M. Staffing intensity, skill mix and mortality outcomes: the case of
chronic obstructive lung disease. Health Services Management Research. 1999;12:258-
268.
119. Hagen S, Hunt J. Investigating the relationship between nurse: patient ratios and patient
outcomes using nationally collected data [abstract]. Abstract Book / Association for Health
Services Research. 1998;15:103.
120. Hunt J, Hagen S. Nurse to patient ratios and patient outcomes. Nurs Times. 1998;94:63-66.
121. Needleman J, Buerhaus PI, Mattke S, Stewart M, Zelevinsky K. Nurse Staffing and Patient
Outcomes in Hospitals. Boston, MA: US Department of Health and Human Services,
Health Resources and Service Administration; 2001.
122. American Nurses Association Network Inc. Implementing nursing's report card: A study
of RN staffing, length of stay, and patient outcomes. Washington, DC: American Nurses
Publishing; 1997:32.
123. American Nurses Association Network Inc. Nurse Staffing and Patient Outcomes in the
Inpatient Hospital Setting. Washington, DC: American Nurses Publishing; 2000:53.
124. Kovner C, Gergen PJ. Nurse staffing levels and adverse events following surgery in U.S.
hospitals. Image J Nurs Sch. 1998;30:315-321.
125. Lichtig LK, Knauf RA, Milholland DK. Some impacts of nursing on acute care hospital
outcomes. J Nurs Adm. 1999;29:25-33.
126. Fridkin SK, Pear SM, Williamson TH, Galgiani JN, Jarvis WR. The role of understaffing
in central venous catheter-associated bloodstream infections. Infect Control Hosp
Epidemiol. 1996;17:150-158.
127. Flood SD, Diers D. Nurse staffing, patient outcome and cost. Nurs Manag . 1988;19:34-35,
38-39, 42-43.
128. Kuhn EM, Hartz AJ, Gottlieb MS, Rimm AA. The relationship of hospital characteristics
and the results of peer review in six large states. Med Care. 1991;29:1028-1038.
129. Wan TT, Shukla RK. Contextual and organizational correlates of the quality of hospital
nursing care. QRB Qual Rev Bull. 1987;13:61-64.
130. Grillo-Peck AM, Risner PB. The Effect of a Partnership Model on Quality and Length of
Stay. Nursing Economics. 1995;13:367-374.
131. Pierce SF. Nurse-sensitive health care outcomes in acute care settings: an integrative
analysis of the literature. Journal of Nursing Care Quality; 1997:60-72.
132. Czaplinski C, Diers D. The effect of staff nursing on length of stay and mortality. Med
Care. 1998;36:1626-1638.
133. Aiken LH, Sloane DM. Effects of specialization and client differentiation on the status of
nurses: the case of AIDS. J Health Soc Behav. 1997;38:203-222.
134. Aubert RE, Herman WH, Waters J, et al. Nurse case management to improve glycemic
control in diabetic patients in a health maintenance organization. A randomized, controlled
trial. Ann Intern Med. 1998;129:605-612.
135. Borkowski V. Implementation of a managed care model in an acute care setting. J Healthc
Qual. 1994;16:25-27, 30.
136. Clark JM, Maben J. Health promotion in primary health care nursing: the development of
quality indicators. Health Educ J. 1999;58:99-119.
137. Dring R, Hiott B, Elliott K. Case management: a case study. J Neurosci Nurs.
1994;26:166-169.
444

138. Jones J, Rowan K. Is there a relationship between the volume of work carried out in
intensive care and its outcome? Int J Technol Assess Health Care. 1995;11:762-769.
139. Strzalka A, Havens DS. Nursing care quality: comparison of unit-hired, hospital float pool,
and agency nurses. J Nurs Care Qual. 1996;10:59-65.
140. Westcott E, Dunn V. An exploration of the value of specialist neurosurgical nurses...
including commentary by Northcott N. NT Research. 1998;3:421-431.
141. Gilbert M, Counsell C. Planned change to implement a restraint reduction program J Nurs
Care Qual. 1999;13:57-64. Published erratum appears in J Nurs Care Qual 1999;13.
142. von Koch L, Widen Holmqvist L, Kostulas V, Almazan J, de Pedro-Cuesta J. A
randomized controlled trial of rehabilitation at home after stroke in Southwest Stockholm:
outcome at six months. Scand J Rehabil Med. 2000;32:80-86.
143. Wagner EH, LaCroix AZ, Grothaus L, et al. Preventing disability and falls in older adults:
a population-based randomized trial. Am J Public Health. 1994;84:1800-1806.
144. Day G, Hindmarsh J, Hojna C, Roy G, Ventimiglia N. Improving medication
administration through an enhanced occurrence reporting system. J Nurs Care Qual.
1994;9:51-56.
145. Mendelson MH, Short LJ, Schechter CB, et al. Study of a needleless intermittent
intravenous-access system for peripheral infusions: analysis of staff, patient, and
institutional outcomes. Infect Control Hosp Epidemiol. 1998;19:401-406.
146. Rush J, Fiorino-Chiovitti R, Kaufman K, Mitchell A. A randomized controlled trial of a
nursery ritual: wearing cover gowns to care for healthy newborns. Birth. 1990;17:25-30.
147. Gilbert B. Facts, not myths. Study shows for-profit hospital conversions help communities
by injecting capital. Mod Healthc; 1997:54.
148. Moore MM, Nguyen D, Nolan SP, et al. Interventions to reduce decibel levels on patient
care units. Am Surg. 1998;64:894-899.
149. Spetz J. Hospital use of nursing personnel: holding steady through the 1990s. J Nurs Adm.
2000;30:344-346.
150. Coffman J, Spetz J. Maintaining an adequate supply of RNs in California. Image J Nurs
Sch. 1999;31:389-393.
151. Coffman J, Spetz J, Seago JA, Rosenoff E, O;Neil E. Nursing in California: A Workforce
in Crisis. . San Francisco, CA: UCSF Center for the Health Professions; 2001:91.
152. Aiken LH, Smith HL, Lake ET. Lower medicare mortality among a set of hospitals known
for good nursing care. Med Care. 1994;32:771-787.
153. Aiken LH, Clarke SP, Sloane DM. Hospital Restructuring: Does it Adversely Affect Care
and Outcomes? J Health Hum Serv Adm. 2000; In press.
154. Archibald LK, Manning ML, Bell LM, Banerjee S, Jarvis WR. Patient density, nurse-topatient
ratio and nosocomial infection risk in a pediatric cardiac intensive care unit. Pediatr
Infect Dis J. 1997;16:1045-1048.
155. Harbarth S, Sudre P, Dharan S, Cadenas M, Pittet D. Outbreak of Enterobacter cloacae
related to understaffing, overcrowding, and poor hygiene practices. Infect Control Hosp
Epidemiol. 1999;20:598-603.
156. Robert J, Fridkin SK, Blumberg HM, et al. The influence of the composition of the nursing
staff on primary bloodstream infection rates in a surgical intensive care unit. Infect Control
Hosp Epidemiol. 2000;21:12-17.
157. Taunton RL, Kleinbeck SV, Stafford R, Woods CQ, Bott MJ. Patient outcomes. Are they
linked to registered nurse absenteeism, separation, or work load? J Nurs Adm. 1994;24:48-
55.
445

158. Amaravadi RK, Dimick JB, Pronovost PJ, Lipsett PA. ICU nurse-to-patient ratio is
associated with complications and resource use after esophagectomy. Intensive Care Med.
2000;26:1857-1862.
159. Silber JH, Rosenbaum PR, Ross RN. Comparing the contributions of groups of predictors:
which outcomes vary with hospital rather than patient characteristics? Journalof the
American Statistical Association. 1995;90:7-18.
160. Zimmerman JE, Shortell SM, Knaus WA, et al. Value and cost of teaching hospitals: a
prospective, multicenter, inception cohort study . Crit Care Med. 1993;21:1432-1442.
161. Wan TT. Hospital variations in adverse patient outcomes. Qual Assur Util Rev. 1992;7:50-
53.
162. Shukla R. Structure vs. people in primary nursing: An inquiry. Nursing Research.
1981;30:236-241.
446

Chapter 40. Promoting a Culture of Safety
Laura T. Pizzi, PharmD
Neil I. Goldfarb
David B. Nash, MD, MBA
Thomas Jefferson University School of Medicine
and Office of Health Policy & Clinical Outcomes
Background
In a number of high hazard organizations, where the risk of error involves dire
consequences, leaders manage for safe, reliable performance. As a result, the term High
Reliability Organization has been coined to describe organizations with exemplary track records
of safety: aviation, chemical manufacturing, shipping, nuclear power production, and the
military. 1-10 This concept is rooted in the analyses of errors that reveal organizational failures,
along with technical failures (related to system performance) and human limitations (related to
human behavior). 11
Theories about antecedents to accidents abound but major schools of thought include
Reason’s belief that a number of latent factors embedded in organizational systems can align and
result in accidents, 12-14 and Rasmussen’s approach to categorizing the different sources of error
that interact with latent factors to produce accidents. 15-17 Another school of thought, developed
by Charles Perrow and first publicized shortly after the Three Mile Island nuclear accident,
Normal Accident Theory, 18,19 emphasizes the ever-present possibility of accidents in
organizations that exhibit complexity and “tight coupling” of processes and the inevitability of
accidents. Normal Accident Theory stands somewhat in opposition to High Reliability Theory,
which holds that accidents can be prevented through organizational design and management.
Scott Sagan’s analysis of the nuclear weapons industry, addressing the question of why there has
never been an ‘accidental’ nuclear war, represents a fascinating investigation of a test case for
these two schools of thought. 20 Despite the obvious apparent confirmation of the High Reliability
Theory perspective (ie, such an accident has thankfully never occurred), Sagan uncovers a
surprising amount of evidence that also seems to confirm the Normal Accident perspective.*
Regardless of the underlying theory, health care is vulnerable to error. The application of
safety promotion theories utilized to positive effect in other high hazard organizations are being
considered for health care, where “accidents” tend to occur one person at a time instead of in
sweeping disasters. 25
Attention to organizational issues of structure, strategy and culture may be a promising
direction for medicine. Although organizational elements are intertwined and must be aligned for
optimum performance 26 this chapter focuses on the culture component, especially “safety
cultures.” Following a description of the prevailing models of culture and safety, we review
approaches that both medical and non-medical industries have used to promote a culture of
safety. On the medical side, the discussion is limited to the Veterans Health Administration’s
comprehensive safety initiative. 27 On the non-medical side, specific methods other high
* Invited commentaries on Diane Vaughn’s in-depth analysis of the Challenger crash 21 also
provide an interesting comparison between the Normal Accidents 22 and High Reliability Theory
23,24 perspectives and indicates that they are more complementary than contradictory.
447

eliability industries have applied to promote a safety culture, 4 including a behavior-based
industry approach, are reported. 28
Organizational Culture
Helmreich defines culture as “a complex framework of national, organizational, and
professional attitudes and values within which groups and individuals function.” 29 Corporate
culture is often referred to as the glue that holds an organization together, and is therefore
assumed to be a contributor to organizational performance by socializing workers in a way that
increases commitment to the goals of the entity. 4,30,31 As such, it embodies the philosophy of
senior leaders, which is translated into, and affects the behaviors of employees. 32 Although some
schools of thought focus on the role of leaders of an organization (board members and
executives), others note that middle management likely plays a substantial role as well,
conveying the culture to front-line workers in any organization, as evidenced by studies of the
effective use of total quality management. 33 The power of culture often goes unrecognized, since
employees may assume that the dominant paradigm is simply “the way we do things here.” 29
Safety Culture
While an exact definition of a safety culture does not exist, a recurring theme in the
literature is that organizations with effective safety cultures share a constant commitment to
safety as a top-level priority, which permeates the entire organization. More concretely, noted
components include: 1) acknowledgment of the high risk, error-prone nature of an organization’s
activities, 2) blame-free environment where individuals are able to report errors or close calls
without punishment, 3) expectation of collaboration across ranks to seek solutions to
vulnerabilities, and 4) willingness on the part of the organization to direct resources to address
safety concerns. 3,4,29,34-36 Based on extensive field work in multiple organizations, Roberts et al
have observed several common, cultural values in reliability enhancing organizations:
“interpersonal responsibility; person centeredness; [co-workers] helpful and supportive of one
another; friendly, open sensitive personal relations; creativity; achieving goals, strong feelings of
credibility; strong feelings of interpersonal trust; and resiliency.” 4
Culture Surveys
The aspect of organizational safety culture that may be visible or measurable is
sometimes referred to as the safety “climate,” which includes management systems, safety
systems, and individual attitudes and perceptions. 32 Health care organizations are now adapting
safety culture and climate surveys from other industries to benchmark and identify potential
deficiencies in their unique safety culture. Kaiser Permanente, the oldest and largest not-forprofit
health maintenance organization in the United States, has administered an executive
attitudes and beliefs survey to identify perceptions of patient safety for the purposes of planning
and measurement (written communication, February 2001, Suzanne Graham). The VA Palo Alto
Patient Safety Center of Inquiry and Stanford University’s Center for Health Policy/Center for
Primary Care and Outcomes Research are conducting a patient safety culture survey that builds
on past work by Gaba and collaborators. The survey includes items on production pressures and
safety consequences, and draws from several other sources (personal communication, June,
2001, Sara Singer). Spath provides a checklist of elements that health care managers can use to
identify which cultural elements should be addressed in order to improve safety 37 (Table 40.1).
Previous work in assessing organizational culture effects on total quality management, 38 and
448

organizational culture in high reliability organizations 39 may also be pertinent to efforts to
measure culture and its consequences for patient safety.
Industries Outside Medicine
Promoting a culture of safety has historically been a priority for the chemical, electrical,
food processing, petroleum, plastic, and transportation industries. Since the 1930s, safety
managers within various industries have recognized that most occupational injuries have a strong
behavioral component, typically rooted in the safety culture. 28 In these settings, behavior
analysis has been used as an approach to solving safety problems. Behavioral analyses typically
involve assessing upstream and downstream behaviors associated with the problem, with further
analysis as to which behaviors may be modifiable. Once relevant behaviors are identified, a
behavior change intervention is implemented, and behavioral changes are measured.
Interventions are customized, and draw upon techniques of behavior science, organizational
development, safety science, and quality. Researchers have shown associations between
behavior-based safety programs and reduced rates of accidents.
In an extensive field study of three organizations (nuclear aircraft carriers, a nuclear
power plant, and the Federal agency responsible for air traffic control) whose operations have
the potential for widespread harm, Roberts et al proposed several management processes that
“cradle” a culture of perfection. 4 One process requires distributing decision making, while
having mechanisms that allow decisions to migrate up and down the chain of command as
circumstances develop. The mechanism for localizing decision making is often extensive
training, while the approach to moving decisions to higher levels is based on management by
exception when acceptable operation is in question. Finally, these researchers suggest that both
top-level managers and local operators develop a deep understanding of their organizations, and
use this “big picture” perspective to provide intuitive judgments when situations arise.
Practice Description
Veterans Health Administration Approach
The Veterans Health Administration (VHA) has implemented a multifaceted safety
initiative, which was designed to build a culture of safety and address system failures. 27 The
approach consists of 4 major elements: 1) partnering with other safety-related organizations and
affiliates to demonstrate a public commitment by leadership, 2) establishing centers to direct
safety efforts, 3) improving reporting systems, and 4) providing incentives to health care team
members and division leaders. These tactics are detailed below. In addition, several specific
initiatives were implemented to address problems, such as bar coding of medications
(Subchapter 43.1) and use of computerized medical records.
To demonstrate a public commitment to the importance of patient safety, the VHA
leadership founded the National Patient Safety Partnership, along with several major healthrelated
organizations (the American Association of Medical Colleges, the American Hospital
Association, the American Medical Association, the American Nurses Association, and the
Institute for Healthcare Improvement). In addition, key senior management officials sounded the
safety message in congressional testimony.
The second part of the VHA's approach involved establishing centers dedicated to the
promotion of patient safety. The first of these was the National Center for Patient Safety, which
directs patient safety efforts for the VHA at a national level. The Director of the Center oversees
patient safety efforts for the entire VHA health system and is a recognized authority.
449

Subsequently, four Patient Safety Centers for Inquiry were funded, which are primarily
responsible for safety-related research and development. Specifically, the centers are responsible
for identifying problems in the patient care process, implementing corrective measures, and
studying effects. Currently, one of these centers is studying safety cultures in health care
organizations. Finally, the VHA's Virtual Learning Center contributes to the safety initiative by
allowing VHA facilities to share lessons learned. Additional information, such as training,
educational programs, alerts, and advisories are planned.
The third major component of the VHA's initiative involves incentives aimed at
improving safety. There are two types of incentives offered: 1) the "carrot," which is a monetary
award of up to $5000 for individuals and teams that develop approaches to improve safety
issues; and 2) the "stick," which is a performance expectation imposed on leaders to improve
patient safety. Leaders of the VHA's 22 regional networks must demonstrate involvement in
safety-promoting activities, or be subject to consequences, including possible termination of
employment. The primary objective of this incentive is to align regional and national leaders'
goals.
Last, the VHA has implemented a two-pronged system for capturing adverse events. The
first of these systems, the Patient Safety Event Registry, mandates reporting of adverse events
and “close calls” occurring within the system. Before implementing the Patient Safety Event
Registry, regional review of event cases was sporadic. After implementation, event data is
systematically shared both regionally and nationally. The second of the systems, the voluntary
reporter identity system, was developed in conjunction with the National Aeronautics and Space
Administration, and allows for anonymous event reporting. It is intended that the use of both
reporting systems will together provide a more comprehensive picture of safety management
than would be possible with one system alone.
Behavior-Based Safety Programs
Outside of medicine, the objective of behavior-based safety interventions is to reduce
incidents by managing at-risk behaviors of the organization and work teams. An approach
described by Krause and colleagues consisted of safety assessments, steering committee
formation, development of checklists of well-specified critical behaviors related to safety,
observer training regarding the critical behaviors, observation and feedback. 28 These steps,
somewhat analogous to aspects of crew resource management training approaches (see Chapter
44), most likely reflect an active safety culture. The Krause study assessed the effectiveness of
behavioral safety initiatives in reducing accidents in 229 facilities in various industries, including
chemical, electrical, food, plastic, petroleum, transportation, service, and paper manufacturers. 28
The study used an interrupted time series design with the participating industrial sites. Event
rates after implementation of the behavioral program were compared with the Occupational
Safety and Health Administration (OSHA) recordable illness/injury rates. Of the 229
participating sites, 73 provided necessary data (others were excluded either because they failed
to provide OSHA illness/injury rates or results of the behavioral initiative). Compared with
baseline, the behavioral initiative resulted in an average 26% improvement in targeted safety
behaviors during the first year, which rose to 69% by the fifth year.
Prevalence and Severity of the Target Safety Problem
There is no known information about the prevalence of medical error emanating from
cultural/organizational problems in health care. Culture is known to contribute to the occurrence
of errors and accidents. Its contribution relative to other causal factors is unknown, but likely to
450

vary, depending on the type of accident and work environment. 3,7,29 The aviation industry
attributes its successful safety record in part to analysis of near miss and accident reports (see
Chapter 4). 40-43 These types of analyses are only possible if the culture supports reporting of
errors. Culture changes may, in fact, have their greatest impact on “underground” (unreported)
errors, which are extremely difficult to quantify.
Opportunities for Impact
Although no data from ongoing surveys has yet emerged to permit us to accurately
quantify safety culture penetration, we nonetheless speculate based on anecdotal evidence that
health care organizations have plenty of room for improvement. A number of observers have
noted large-scale obstacles to promotion of safety culture within health care: a pervasive culture
of blame that impedes acknowledgment of error, and professional “silos” that offer unique
challenges to changing any universal aspect of health care, including culture. 44-46
Even before the Institute of Medicine’s pivotal To Err is Human report was delivered to
the public, promoting a safety culture within health care had received widespread attention. The
Institute for Healthcare Improvement’s Web site features a report “Reducing Medical Errors and
Improving Patient Safety: Success Stories from the Front Lines of Medicine.” 47 It includes
articles about the transformation of culture at the prestigious Dana-Farber Cancer Institute after a
highly publicized chemotherapy overdose in 1994, which resulted in the death of a patient.
Another article in the same series highlighted the major steps, including cultural change, as noted
above, taken by leaders of the nation’s largest health care provider—the Veterans Affairs
Healthcare System—after fatal medical errors were reported by the media. 47
Comment
Measuring the impact of culture on safety-related outcomes is challenging. Culture is a
complex and abstract construct that must be inferred from behaviors, and analysis often relies on
self-reported data. 29 Research continues to develop a working model of safety culture that
permits measurement of several connected concepts: individuals’ perceptions and attitudes about
safety, individuals’ observable safety behaviors, and an organization’s safety management
system as evidenced by its policies and management styles. 35 The relative impact of each of
these measures on outcomes is another layer of ongoing research.
Although some data support the effectiveness of the entire VHA initiative in improving
safety, there are no direct data supporting the effect of promoting a culture of safety. The use of
incentives to reward safety-promoting behavior and publicly demonstrating a commitment to
safety are approaches that could be applied in both large and small health care settings. The
VHA’s reporting system will likely be watched, and potentially adapted by large providers who
have inconsistent and/or insufficient reporting of safety problems at local, regional, and national
levels.
The evidence presented by Krause provides compelling support for the effectiveness of
behavior-based safety programs in a wide range of industrial settings. Although this exact
approach has not been evaluated in health care environments, its emphasis on promoting safety
culture seems applicable to patient care environments.
As noted in To Err is Human, researchers who have studied organizations with a strong
safety culture believe that it is “the most critical underlying feature of their accomplishments.” 48
Although the nature of the evidence is based on field studies and other methods not typical of
medical evidence, it is considered compelling by a number of experts from organizational and
other social sciences. At this point, promoting a culture of safety remains surprisingly
451

unexplored in health care settings, where the risks of error are high. Further research in this area
is warranted, though the threshold for evidence may need a different yardstick than is typically
applied in medicine (Chapter 2).
Acknowledgments
The authors are grateful to Mr. Marcus Moore of Behavioral Science Technology, Inc.
for providing information related to industrial safety programs. We would also like to thank Dr.
Kasey Thompson, Director of the American Society of Health-System Pharmacy’s Center for
Patient Safety, and Mr. Mark Thomas, Director Of Pharmacy at Children’s Hospital in
Minneapolis for their suggestions and insight on cultural safety issues.
452

Table 40.1. Checklist of elements that contribute to a patient-safe environment
‰ All people acknowledge that top management provides essential patient safety improvement leadership.
‰ The organization has clearly defined patient safety polices.
‰ All people can explain the organization’s patient safety policies.
‰ All people are involved in developing patient safety goals, and everyone can explain desired
results and measures.
‰ All people are actively involved in identifying and resolving patient safety concerns.
‰ All people can explain how their personal performance affects patient safety.
‰ All people believe they have the necessary authority and resources to meet their
responsibilities for patient safety
‰ Patient safety performance for all people is measured against goal, clearly displayed, and
rewarded.
‰ A comprehensive review of patient safety is conducted annually, and there is a process in
place that drives continuous improvement.
‰ Regular workplace hazard analyses are conducted to identify patient safety improvement
opportunities. The results are used to make changes in patient care activities.
‰ All people are empowered to correct patient safety hazards as they are identified.
‰ A comprehensive system exists for gathering information on patient safety hazards. The
system is positive, rewarding, and effective, and people use it.
‰ All people are fully aware of patient incident trends, causes, and means of prevention.
‰ All injury-producing patient incidents and significant “near misses” are investigated for root
cause, with effective preventive actions taken.
‰ All people who operate patient care equipment are trained to recognize maintenance needs
and perform or request timely maintenance.
‰ All people know immediately how to respond to an emergency because of effective planning,
training, and drills.
‰ Facilities are fully equipped for emergencies; all necessary systems and equipment are in
place and regularly tested; and all people know how to use equipment and communicate
during emergencies.
‰ Ergonomics experts are provided when needed and are involved in patient safety assessment
and training.
453

‰ All supervisors/managers assist in patient safety workplace analyses, ensure physical
protections, reinforce training, enforce discipline, and can explain how to provide safe
patient care.
454

References
1. Roberts KH, Gargano G. Managing a high reliability organization: A case for
interdependence. In: Glinow MAV, Mohrmon S, editors. Managing complexity in high
technology industries: Systems and people. New York: Oxford University Press.; 1989:
147-59.
2. Gaba DM. Structural and organizational issues in patient safety: A comparison of health
care to other high-hazard industries. California Management Review. 2000; 43: 83-102.
3. Roberts KH. Some characteristics of high reliability organizations. Berkeley, CA ::
Produced and distributed by Center for Research in Management, University of California,
Berkeley Business School; 1988.
4. Roberts KH. Cultural characteristics of reliability enhancing organizations. Journal of
Managerial Issues. 1993; 5: 165-81.
5. Roberts KH. Managing high reliability organizations. California Management Review.
1990; 32: 101-13.
6. Roberts KH, Stout SK, Halpern JJ. Decision dynamics in two high reliability military
organizations. Management Science. 1994; 40: 614-24.
7. Weick KE. Organizational culture as a source of high reliability. California management
review. 1987; 29: 112-27.
8. Roberts KH, Libuser C. From Bhopal to banking: Organizational design can mitigate risk.
Organizational Dynamics. 1993; 21: 15-26.
9. LaPorte TR, Consolini P. Theoretical and operational challenges of "high-reliability
organizations": Air-traffic control and aircraft carriers. International Journal of Public
Administration. 1998; 21: 847-52.
10. LaPorte TR. The United States air traffic control system : increasing reliability in the midst
of rapid growth. In: Mayntz R, Hughes TP, editors. The Development of Large technical
Systems. Boulder, CO:: Westview Press; 1988.
11. Van Vuuren W. Organizational failure: lessons from industry applied to the medical
domain. Safety Science. 1999; 33: 14-29.
12. Reason JT. Human Error. New York: Cambridge Univ Press; 1990.
13. Reason JT. Managing the Risks of Organizational Accidents: Ashgate Publishing
Company; 1997.
14. Reason J. Human error: models and management. BMJ. 2000; 320: 768-70.
15. Rasmussen J. Human error and the problem of causality in analysis of accidents. Philos
Trans R Soc Lond B Biol Sci. 1990; 327: 449-60.
16. Rasmussen J. Afterword. In: Bogner MS, editor. Human error in medicine. Hillsdale, N.J.:
L. Erlbaum Associates; 1994: 385-93.
17. Rasmussen J, Goodstein LP, Andersen HB, Olsen SE. Tasks, errors, and mental models : a
festschrift to celebrate the 60th birthday of Professor Jens Rasmussen. London ; New York
:: Taylor & Francis,; 1988.
18. Perrow C, Langton J. The limits of safety: the enhancement of a theory of accidents.
Journal of Contigency Management. 1994; 2: 212-20.
19. Perrow C. Normal accidents : Living with High-Risk Technologies. With a New Afterword
and a Postscript on the Y2K Problem. Princeton, NJ:: Princeton University Press; 1999.
20. Sagan SD. The Limits of Safety: Organizations, Accidents and Nuclear Weapons.
Princeton, N.J.: Princeton University Press; 1993.
455

21. Vaughn D. The Challenger Launch Decision : Risky Technology, Culture, and Deviance at
Nasa. Chicago, Ill: Univ of Chicago Press; 1997.
22. Sagan SD. The Challenger Launch Decision: Risky Technology, Culture, and Deviance at
NASA. Administrative Science Quarterly. 1997; 42: 401-05.
23. Roberts KH. The Challenger Launch Decision: Risky Technology, Culture, and Deviance
at NASA. Administrative Science Quarterly. 1997; 42: 405-10.
24. Weick KE. The Challenger Launch Decision: Risky Technology, Culture, and Deviance at
NASA. Administrative Science Quarterly. 1997; 42: 395-401.
25. Gaba DM. Structural and organizational issues in patient safety: a comparison of health
care to other high-hazard industries. California Management Review. 2000; 43: 1-20.
26. O'Brien JL, Shortell SM, Hughes EFX, Foster RW, Carman JM, Boerstler H, O'Connor EJ.
An Integrative Model for Organization-wide Quality Improvement: Lessons from the Field.
Quality Management in Health Care. 1995, 3:19-30.
27. Weeks WB, Bagian JP. Developing a culture of safety in the Veterans Health
Administration. Eff Clin Pract. 2000; 3: 270-6.
28. Krause TR, Seymour KJ, Sloat KCM. Long-term evaluation of a behavior-based method
for improving safety performance: a meta-analysis of 73 interrupted time-series
replications. Safety Science. 1999; 32: 1-18.
29. Helmreich RL, Merritt AC. Culture at work in aviation and medicine : national,
organizational, and professional influences. Aldershot ; Brookfield, VT, USA :: Ashgate;
1998.
30. Peters TJ, Waterman RH. In search of excellence : lessons from America's best-run
companies. New York :: Harper & Row; 1982.
31. Deal TE, Kennedy AA. Corporate cultures : the rites and rituals of corporate life.
Reading, Mass. :: Addison-Wesley Pub. Co.; 1982.
32. Spath P. Does your facility have a 'patient-safe' climate? Hosp Peer Rev. 2000; 25: 80-2.
33. Baskin, Susan T (Osborn) and Shortell SM. Total Quality Management: needed research on
the structural and cultural dimensions of quality improvement in health care organizations.
The Journal of Health Administration Education. 1995; 13:143-154.
34. Helmreich RL, Foushee HC, Benson R, Russini W. Cockpit resource management:
exploring the attitude-performance linkage. Aviat Space Environ Med. 1986; 57: 1198-200.
35. Cooper MD. Towards a model of safety culture. Safety Science. 2000; 36: 111-36.
36. Geller ES. Ten Leadership qualities for a total safety culture. Professional Safety. 2000; 45:
30-32.
37. Spath P. Patient safety improvement guidebook. Forest Grove, OR: Brown-Spath &
Associates; 2000.
38. Shortell SM, Jones RH, Rademaker AW, Gillies RR, Dranove DS, Hughes EF, et al.
Assessing the impact of total quality management and organizational culture on multiple
outcomes of care for coronary artery bypass graft surgery patients. Med Care. 2000; 38:
207-17.
39. Klein RL, Bigley GA, Roberts KH. Organizational culture in high reliability organizations:
An extension. Human Relations. 1995; 48: 771-93.
40. Billings CE, Reynard WD. Human factors in aircraft incidents: results of a 7-year study.
Aviat Space Environ Med. 1984; 55: 960-5.
41. Billings CE. Some hopes and concerns regarding medical event-reporting systems: lessons
from the NASA Aviation Safety Reporting System. Arch Pathol Lab Med. 1998; 122: 214-
5.
456

42. Helmreich RL. On error management: lessons from aviation. BMJ. 2000; 320: 781-85.
43. Barach P, Small SD. Reporting and preventing medical mishaps: lessons from non-medical
near miss reporting systems. BMJ. 2000; 320: 759-63.
44. Umiker W. Organizational culture: the role of management and supervisors. Health Care
Superv. 1999; 17: 22-7.
45. Glavin MPV, Chilingerian JA. Hospital care production and medical errors: organizational
responses to improve care. Current Top Management. 1998; 3: 193-215.
46. Manasee H, Jr. Close to home. Your own staff can solve many problems with your
hospital's medication system. Hosp Health Netw. 2001; 75: 82.
47. Findlay SE National Coalition on Health Care and the Institute for Healthcare
Improvement. Reducing medical errors and improving patient safety. Available at:
http://www.ihi.org/resources/act/medical_errors.pdf. Accessed June 26, 2001.
48. Kohn L, Corrigan J, Donaldson M, editors. To Err Is Human: Building a Safer Health
System. Washington, DC: Committee on Quality of Health Care in America, Institute of
Medicine. National Academy Press; 2000.
457

Section G. Systems Issues and Human Factors
Chapter 41. Human Factors and Medical Devices
Subchapter 41.1. The Use of Human Factors in Reducing Device-related
Medical Errors
Subchapter 41.2. Refining the Performance of Medical Device Alarms
Subchapter 41.3. Equipment Checklists in Anesthesia
Chapter 42. Information Transfer
Subchapter 42.1. Information Transfer Between Inpatient and Outpatient
Pharmacies
Subchapter 42.2. Sign-Out Systems for Cross-Coverage
Subchapter 42.3. Discharge Summaries and Follow-up
Subchapter 42.4. Notifying Patients of Abnormal Results
Chapter 43. Prevention of Misidentifications
Subchapter 43.1. Bar Coding
Subchapter 43.2. Strategies to Avoid Wrong-Site Surgery
Chapter 44. Crew Resource Management and its Applications in Medicine
Chapter 45. Simulator-Based Training and Patient Safety
Chapter 46. Fatigue, Sleepiness, and Medical Errors
Chapter 47. Safety During Transportation of Critically Ill Patients
Subchapter 47.1. Interhospital Transport
Subchapter 47.2. Intrahospital Transport
458

Chapter 41. Human Factors and Medical Devices
Harvey J. Murff, MD
Harvard Medical School
John W. Gosbee, MD, MS
Deparment of Veterans Affairs National Center for Patient Safety
David W. Bates, MD, MSc
Harvard Medical School
Introduction
Human factors engineering (HFE), also known as usability engineering or ergonomics, is
the study of how humans interact with machines and complex systems. 1 Through the merging of
cognitive psychology, engineering and other disciplines, human factors researchers have detailed
numerous principles concerning device and software program designs that allow for optimal
usage. 2 When these principles are violated, improper use of a machine is more likely to result. 3,4
Specific examples have been detailed during observations of user errors with electronic infusion
devices. 5
Medical device misuse is an important cause of medical error, 6,7 and therefore,
incorporating human factors methodology into the design of medical devices has assumed an
important role in ensuring patient safety. 2,8 This chapter first describes the use of HFE principles
as a safety practice in the design of medical devices and their evaluation both prior to and after
institutional purchase. Next, medical device alarms and the contribution of HFE to alarm
improvements will be evaluated. Finally, the chapter reviews the use of preoperative checklist
procedures to reduce anesthesia device failures (see also Chapter 23).
Subchapter 41.1. The Use of Human Factors in Reducing Device-related Medical Errors
Background
Human factors engineering is a powerful component in the design of usable, safe medical
devices. 8 HFE principles can be incorporated as safety practices that occur at various points
during device development and usage. Industry can use HFE principles at multiple times in the
design and developmental cycle of medical devices and software packages. 3 Health care
institutions can consider results of HFE evaluations when deciding which products to purchase.
Finally, HFE principles can also be incorporated into the ongoing evaluation of devices that have
already been purchased and are in use. While these practices have high face validity, there has
been little formal study of their effectiveness in reducing medical error. They are presented here
because they may hold promise if scrutinized rigorously, and to familiarize readers with their
potential to reduce medical error.
Design and Developmental Phase
Data collected by the United States Food and Drug Administration (FDA) in the late
1980s demonstrated that almost half of all medical device recalls resulted from design flaws. 9 In
1990, Congress passed the Safe Medical Devices Act, giving the FDA the ability to mandate
good manufacturing practices (GMP). These GMP involve design controls for manufacturers
that help ensure the use of HFE within medical device design. 9 As described in the Good
459

Manufacturing Practice Regulation, Design Control subsection (Title 21-Section 820.30), these
include the use of iterative design and testing during the developmental phase. The Act requires
that designs be “appropriate and address the intended use of the device, including the needs of
the user and patient.” 10 Multiple human factors techniques, such as user studies, prototype tests,
and task/function analysis, are utilized in the development and design process.
Manufacturers are required not only to use human factors principles to repeatedly test the
product in all phases of design, but also to validate the ultimate device design. Validation entails
testing the device, either in an actual clinical situation or a simulation, and documenting that the
device conforms to the individual user’s needs. Thus, manufacturers are required to apply HFE
methods through the multiple phases of device design and development cycles. 10
Human factors engineering practices for medical device design and evaluation have been
well described. In 1993 the Association for the Advancement of Medical Instrumentation and the
American National Standards Institute established guidelines for the incorporation of HFE
principles into medical device design. 11 This comprehensive document helped direct attention to
the problem of poor medical device design and helped establish the design standards necessary
to ensure safe medical equipment and thus should stand as the safety benchmark for industry.
Only limited data are available concerning the application of HFE principles to medical
device design, and most are not published. Nonetheless, the application of human factors
principles during a device’s design phase has been demonstrated to reduce user error. Patient
controlled analgesia (PCA) pumps are a case in point of how HFE principles in product design
reduce user error. User errors associated with poor interface design have been described with
PCA pumps. 12,13 Lin and colleagues investigated whether applying human factors engineering
principles to the design of the user interface of a PCA pump could result in fewer dosage errors
as well as less time spent programming the device. 14 Information on device usage was obtained
through cognitive task analysis. This involved observing and interviewing nurses operating PCA
pumps both in the laboratory setting and in the field. Utilizing this feedback, as well as other
human factors design principles, a “new” PCA pump interface was designed. Twelve recovery
room nurses were required to complete specific tasks with both the standard PCA user interface
and the newly designed interface. There were 29 programming errors on the traditional interface
and 13 on the redesigned interface (an error reduction of 55%, p

Device Evaluation Prior to Purchase
Adhering to HFE principles during initial design stages of a medical device is essential.
However, human factors analysis should also be incorporated into the institutional decision to
acquire a new medical device or software program. 3 Device purchasers should strongly consider
institution-specific human factors testing. Usability testing at the institutional level establishes
built-in redundancies to capture any design problems missed by manufacturers. Furthermore, the
users and environments at individual institutions will differ, possibly in important ways, from the
users and environments in which the device or program was initially designed and tested. It is
important for an institution to be aware of who the intended users of the device or software will
be, as well as where and when they plan to use the device. The information for such evaluations
may be obtained from vendors, from an in-house analysis, or from independent organizations.
Vendors must be able to prove to the FDA that the user will be able to operate the
medical device in the way in which it was intended. 10 As companies are required to collect
human factors analysis data, it is important that institutions wishing to purchase a new medical
device or software receive and carefully review this information. Gosbee provides a list of
questions to ask a vendor before a purchase, which include: “How long does it take to learn to
operate the system? How long does it take to complete typical set-up tasks? What are the types
and frequency of errors that could happen, and the systems to thwart them?” 3
It is also important to consider the environment in which a device will be used.
Idiosyncratic features of the environment, such as excessive noise or poor lighting, and
differences in user skill or acuity due to fatigue or otherwise, may affect safety and the device’s
in-house usability.
Some institutions have developed in-house usability labs, in order to rigorously test any
device before purchasing. The Mayo Clinic uses simulations to test the usability of medical
software before purchasing. 17 By carefully measuring user performance with the software they
are able to uncover latent errors in the design. The usability lab is also able to measure the time
necessary to learn to use the new software. This important information can help predict the
device’s or software’s influence on workflow as well as its predilection for operator misuse.
Even without sophisticated usability laboratories, an institution can use basic human
factors techniques to evaluate a product before purchase. 3 Powerful techniques such as cognitive
walk-through can be easily utilized at any institution. This involves observing the end-users of a
product interact with the product. As they attempt to use the device, they are instructed to “think
out loud.” Careful observation of the user’s actions and comments can identify potential design
flaws that might make it difficult to utilize the device or software.
Independent organizations are another potential source of information on device safety.
Unfortunately, most independent sources do not make clear to what degree HFE principles were
used in product evaluations, although they do provide some assessment of safety. One such
organization is ECRI (formerly the Emergency Care Research Institute), a nonprofit international
health services research agency. Another is the Institute of Safe Medical Practices (ISMP). Both
release newsletters and publications regarding product safety. By searching these and similar
databases, institutions can gather additional information concerning product safety prior to
purchasing a device. ERCI also publishes articles specifically geared to the institutions that
might wish to purchase a medical device or software.
Regardless of the level of pre-procurement testing, some unsafe designs will not be
detected until after the product is in use. 3 Therefore, it is important for institutions to
continuously evaluate these products to ensure safety.
461

Ongoing Device Evaluation
Devices and software at greatest risk for user error should be systematically evaluated.
This is particularly important in areas where multiple devices are used with different interfaces,
such as the operating room or the intensive care units. 3 Furthermore, areas where multiple
medications are stored together should be scrutinized for potential latent errors within device or
software user interfaces prior to user errors occurring.
Resources are available that can help direct an institution’s search. Through publications
from the FDA, ECRI, ISMP and similar organizations, medical device problems identified at
other institutions can be targeted. Thus an important safety practice may be using this published
information to search for latent errors within already purchased medical devices and applying
this information toward a directed product evaluation at the local institution.
Another potential safety practice is to educate practitioners about HFE principles to
increase awareness of medical device user error. 3 Several groups, including the American
Nurses’ Credentialing Center and the American Society of Health-Systems Pharmacists,
recommend incorporating HFE training within health care curricula as a means to reduce error. 18
To create a culture of safety within medicine, practitioners must couple the ability to
identify potential design weaknesses with a change in the prevailing culture of silence
surrounding medical errors. Educational programs directed at health care providers in training
should address both of these important concerns. Curricula for teaching medical student and
medical residents HFE principles have been described 18 and will likely be adopted at other
institutions. Casarett and Helms caution that an unintended result of error curriculum 19 may be
that residents become too willing to attribute an error to system causes. Their concern is that the
resident will ignore any possible individual contribution to the adverse medical event and not
learn from analyses of the event. This concern has been discounted by Gosbee, stating that any
error-in-medicine curriculum should aim to “teach residents to see when errors are due to
inadequate skills and knowledge versus when they are due to inherent cognitive limitations and
biases.” 18
Subchapter 41.2. Refining the Performance of Medical Device Alarms
Background
Numerous aspects of patient care compete for providers’ attention and can reduce their
vigilance in monitoring medical devices. Alarms can alert providers to urgent situations that
might have been missed due to other distractions and have become a necessary part of patient
monitoring. In a study looking at critical incidents within a neonatal intensive care unit, 10%
were detected through alarms. 20
However, fundamental flaws in the design of current alarm systems likely decrease their
impact. 21 There are reports documenting some alarm failings in the medical literature, 22 but few
data address interventions to improve alarm system effectiveness. For an alarm to be effective it
requires that a medical problem trigger the alarm, that personnel identify the source and reason
for the alarm, and that the medical problem be corrected prior to patient injury. This section
reviews 2 aspects of alarm safety: (1) the use of HFE principles in the redesign of medical alarms
to improve identification of the source and reason for alarm, and (2) practices in both device
design and programming that may improve safety by decreasing false positive alarms.
462

Identification of Alarm Source and Reason
The recognition accuracy of alarms within the operating room is quite low. When
presented with alarm sounds and asked to identify the source, anesthesiologists, operating room
technicians, and operating room nurses correctly identify the device producing the alarm only 33
to 53.8% of the time. 23-25 Furthermore, experiments suggest that humans have difficulty reliably
recognizing more than 6 alarms at one time. 26 The sheer number of different medical devices
with alarms can make it difficult to discern one alarm from another and studies within the human
factors literature have documented the inability of medical providers to discern between high
priority and low priority alarms. 27 While this is a known problem in operating rooms and
intensive care units, how well alarms are recognized in other settings has not been described.
Some effort has been made to improve alarm systems through redesign. 28 One nonmedical
study examined ways to improve the recognition of auditory alarms by comparing
abstract alarm sounds with specially designed alarms using speech and auditory icons. 29 Other
studies within the human factors literature have revealed certain acoustical properties that are
more likely to result in a higher sense of perceived urgency by the operator.
In a series of experiments, Edworthy required subjects to rank the level of urgency
associated with different alarms. 30 The acoustical properties of the alarms were altered for the
different subjects. Level of urgency was then correlated with a specific alarm sound. After
ranking a set of acoustic parameters based on perceived urgency, the experimenters predicted
what urgency ranking the alarm would receive and played the alarms for a new set of subjects.
The correlation between the subjects’ urgency rating and the investigators’ predicted ratings was
93% (p

cases, with subjects serving as their own controls. In one study, subjects were required to
identify when changes in physiologic parameters occurred using different visual formats. 34
Response time and accuracy to the simulated cases was compared among a histogram, polygon,
and numerical display. Subject responses were more accurate with the histogram and polygon
displays (p=0.01).
In the other study, 20 anesthesiologists with an average working experience of 5 years
were required to perform specific tasks on an anesthesia simulator 35 (see Chapter 45). The tasks
consisted of returning a set of abnormal hemodynamic parameters to normal using intravenous
medications. A specific time for the completion was determined and this time was compared
among 3 different visual interfaces. Trial time was significantly shorter with the traditional
display (p

Decreasing the Frequency of Alarms
Poorly designed device alarms can create not only problems with alarm recognition but
also frequent false positive alarms. Two observational studies found that from 72 to 75% of
alarms during routine general anesthesia did not require corrective action. 36,37 Another study
showed that only 3% of all auditory alarms during routine anesthesia monitoring represented a
patient risk. 38 Providers frequently must interrupt clinical tasks to silence these false positive
alarms. More concerning is the fact that when alarms are unreliable, they tend to be ignored. 21,39
This “cry-wolf” effect is a significant detriment to the optimal performance of alarm systems and
may result in dire consequences when “true alarms” are ignored.
False alarms can be managed in two ways. Devices can be designed so that they identify
and eliminate false alarms before triggering or users can manipulate alarm parameters to reduce
false alarms. User manipulation can range from adjusting alarm thresholds 40 to even turning the
alarms off. 22 There are no data describing how often operators reset alarm parameters to reduce
false positive rates.
Some research has focused on the identification of alarm parameters that improve or
optimize alarm accuracy (ie, to improve the ratio of true positives to false positives—the “signalto-noise”
ratio). For example, Rheineck-Leyssius and Kalkman studied how altering an alarm
parameter on a pulse oximeter would affect the incidence of hypoxemia. 40 Consecutive patients
admitted to the recovery room of a regional hospital in the Netherlands after general or regional
anesthesia were randomized to either a lower limit of SpO2 90% or SpO2 85%. The 2 groups
were comparable at baseline. The outcomes measured were hypoxemia, defined by a pulse
oximeter reading less than or equal to 90% or 85%. The authors were also required to judge if
they believed a signal to be artifact versus a true positive. The authors were blinded as to which
group the subject was randomized to during artifact assessment and data analysis. The relative
risk of having a hypoxic episode (Sp02≤85%) in the group with the lower alarm limit set at 85%
(as compared with those with the lower alarm limit set at 90%) was 3.10 (95% CI: 1.32-7.28,
p

motion-detection testing to the identified pulse. The “smart” pulse-oximeter only triggered one
false positive (an alarm that did not coincide with hypoxia) and had a relative risk of 0.09 (95%
CI: 0.02-0.48) for generating a false positive alarm when compared with conventional pulse
oximetry with a 21-second delay.
Comment
Observational studies have suggested that current alarm systems could be improved, but
future laboratory studies are needed to determine which properties of alarm systems are most
effective to alert operators. These tests must be followed by field studies and ultimately with
trials looking at actual patient outcomes to determine the best designs. Information concerning
the cost and feasibility of implementing these changes should also be gathered.
False positive alarms remain a significant problem. Few data exist on the incidence of
resetting alarm parameters or at what parameter values alarm accuracy is optimized. Advances in
alarm technology aimed at reducing false positives appear a promising alternative to resetting
parameters.
Subchapter 41.3. Equipment Checklists in Anesthesia
Events related to medical equipment can be divided into two categories, user-error and
equipment failure. 43 Health device inspection and preventive maintenance by biomedical or
clinical engineering departments have high face validity as an important patient safety practice in
reducing equipment failure.
There are many calls in the engineering literature to standardize equipment
maintenance. 44-46,46 Standardization of protocols is believed to help make the processes more
efficient and reduce errors. 47 However, it has been difficult to standardize equipment
maintenance practices due to a lack of the appropriate units on which to base measurement. 46
Some authorities have suggested outcomes based on engineering endpoints such as reliability
and accuracy. 48 Others have tried to validate a set of maintenance outcome units based on cost or
quality metrics. 44,45,49 Some engineers have suggested the incorporation of clinical endpoints into
medical equipment assessment. 48,50 Notwithstanding differing views as to measurement of
endpoints, experts uniformly believe that standardization of engineering endpoints is vital to
ensure adequately inspected and maintained equipment. 46 No studies to date have developed a
widely used standardized protocol for equipment maintenance for clinical engineering
departments, largely because the lack of standardization of endpoints renders assessing the
relative value of any particular maintenance protocol impossible. 44-46,48,50 Nonetheless,
equipment failure does result in a small fraction of clinical events and thus is an important safety
intervention. Hopefully, future studies will help delineate the most effective practices for
equipment maintenance processes.
Use of checklists is another practice that helps ensure equipment readiness, particularly
for equipment that is needed in critical situations and/or where equipment failure may have dire
consequences. For example, a nurse at the beginning of each shift may use a checklist to ensure
the readiness of a hospital ward’s resuscitation cart (“crash cart”) should it be needed (eg, the
defibrillator is plugged-in and charged, the back-up suction pump works, medication is not past
its expiration date). Similarly, a perfusion technologist can use a checklist to ensure cardiac
bypass circuit and back-up equipment are ready before surgery. Published studies on the
effectiveness of equipment checklists largely relate to the use of preoperative checklists to
prevent anesthesia equipment failures since, to date, studies on the effectiveness of equipment
466

checklists in medicine have been limited to this area. 51-53,54 These studies are reviewed in
Chapter 23.
Final Comment to Chapter 41
Human factors testing is yielding important data regarding safe and effective medical
device and alarm designs that take into account the users’ cognitive limitations. Machines can be
designed and redesigned that enhance patient safety, rather than compromise it.
Currently, there are no widely accepted standards for equipment maintenance intervals
and protocols. Maintenance endpoints that incorporate clinical events as one component of the
endpoint have been suggested. Until a reliable and validated engineering endpoint metric is
widely recognized it will remain difficult to investigate the most effective maintenance practices.
Other than the pioneering work in anesthesiology, HFE has been underutilized in
medicine. Hopefully, in the near future, more attention will be focused on integrating human
factors engineering within all aspects of medical training and practice, which will help create a
culture of safety.
References
1. Handbook of Human Factors and Ergonomics. 2 ed. New York: John Wiley and Sons,
INC, 1997.
2. Sawyer D. Do It By Design: An Introduction to Human Factors in Medical Devices. 1997.
FDA.
3. Gosbee J, Lin L. The Role of Human Factors Engineering in Medical Device and Medical
Systems Errors. In: Vincent C, editor. Clinical Risk Management: Enhancing Patient
Safety. 2000.
4. Wiklund M, Gardner-Bonneau D, Carstensen P, Weinger M. Global standardization of
human factors for medical devices and systems. In: Proceedings of the Human Factors and
Ergonomics Society 44th Annual Meeting. Santa Barbara, CA: Human Factors and
Ergonomics Society, 2000: 533-536.
5. Cook RI, Woods DD, Howie MB, Horrow JC, Gaba DM. Case 2-1992. Unintentional
delivery of vasoactive drugs with an electromechanical infusion device. J Cardiothorac
Vasc Anesth. 1992; 6:238-244.
6. Cooper JB, Newbower RS, Kitz RJ. An analysis of major errors and equipment failures in
anesthesia management: considerations for prevention and detection. Anesthesiology. 1984;
60:34-42.
7. Ludbrook GL, Webb RK, Fox MA, Singleton RJ. The Australian Incident Monitoring
Study. Problems before induction of anaesthesia: an analysis of 2000 incident reports.
Anaesth Intensive Care. 1993; 21:593-595.
8. Bogner MS. Designing medical devices to reduce the likelihood of error. Biomed Instrum
Technol. 1999; 33:108-113.
9. Human Factors Implications of the New GMP Rule.
http://www.fda.gov/cdrh/humfac/hufacimp.html . 4-22-1998.
10. Medical Devices; Current Good Manufacturing Practice (CGMP) Final Rule; Quality
System Regulation. http://www.fda.gov/cdrh/humfac/frqsr.html 61[195], 52601-52662. 10-
7-1996. Food and Drug Administration.
11. American National Standards Institute AftAoMI. Human factors engineering guidelines
and preferred practices for the design of medical devices (ANSI/AAMI HE-48). Arlington,
VA: Association for the Advancement of Medical Instrumentation, 1993.
467

12. Brown SL, Bogner MS, Parmentier CM, Taylor JB. Human Error and Patient-Controlled
Analgesia Pumps. Journal of Intravenous Nursing. 1997; 20:311-316.
13. Lin L. Human Error in Patient-Controlled Analgesia: Incident Reports and Experimental
Evaluation. In: Proceedings of the Human Factors and Ergonomics Society 42nd Annual
Meeting. Santa Barbara, CA: Human Factors and Ergonomics Society, 1998: 1043-1047.
14. Lin L, Isla R, Doniz K, Harkness H, Vicente KJ, Doyle DJ. Applying human factors to the
design of medical equipment: patient- controlled analgesia. J Clin Monit Comput. 1998;
14(4):253-263.
15. Aucella A, Kirkham T, Barnhart S, Murphy L, LaConte K. Improving ultrasound systems
by user-centered design. In: Proceedings of the Human Factors and Ergonomics Society.
Santa Barbara, CA: Human Factors and Ergonomics Society, 1994.
16. Sawyer D. Medical device requirements, human factors, and the food and drug
administration. In: Proceedings of the Human Factors and Ergonomics Society 44th
Annual Meeting. Santa Barbara: Human Factors and Ergonomics Society, 2000: 526-527.
17. Claus P, Gibbons P, Kaihoi, BH, Mathiowetz M. Usability lab: A new tool for process
anaylsis at the Mayo Clinic. In: HIMSS Proceedings. Chicago: Healthcare Information
Management Systems Society, 1997: 149-159.
18. Gosbee J. Human Factors Engineering is the Basis for a Practical Error-in-Medicine
Curriculum. First Workshop on Human Error in Clinical Systems.
http://www.dcs.gla.ac.uk/~johnson/papers/HECS_99/Gosbee.html . 1999.
19. Casarett D, Helms C. Systems error versus physicians' errors: Finding the balance in
medical education. Acad Med. 1999; 74:19-22.
20. Frey B, Kehrer B, Losa M, Braun H, Berweger L, Micallef J et al. Comprehensive critical
incident monitoring in a neonatal-pediatric intensive care unit: experience with the system
approach. Intensive Care Med. 2000; 26(1):69-74.
21. Woods D. The Alarm Problem and Direct Attention in Dynamic Fault Management.
Ergonomics. 1995; 38(11):2371-2393.
22. Samuels SI. An alarming problem. Anesthesiology. 1986; 64(1):128.
23. Finley GA, Cohen AJ. Perceived urgency and the anaesthetist: responses to common
operating room monitor alarms. Can J Anaesth. 1991; 38(8):958-964.
24. Loeb RG, Jones BR, Leonard RA, Behrman K. Recognition accuracy of current operating
room alarms. Anesth Analg. 1992; 75(4):499-505.
25. Momtahan K, Hetu R, Tansley B. Audibility and identification of auditory alarms in the
operating room and intensive care unit. Ergonomics. 1993; 36:1159-1176.
26. Patterson R, Milroy R. Auditory warnings on civil aircraft: the learning and retention of
warnings. Final Contract Report 7D/S/0142. 1980. Cambridge, MRC Applied Psychology
Unit.
27. Simons D, Frederiks T, Tappel J. The evaluation of an auditory alarm for a new medical
device. In: Proceedings of the Human Factors and Ergonomics Society 41st Annual
Meeting. Santa Barbara: Human Factors and Ergonomics Society, 1997: 777-781.
28. Edworthy J, Stanton N. A user-centred approach to the design and evaluation of auditory
warning signals: 1. Methodology. Ergonomics. 1995; 38:2262-2280.
29. Leung YK, Smith S, Parker S, Martin R. Learning and Retentiono of Auditory Warnings.
www.icad.org/websiteV2.0/conferences/icad97/leung.pdf . 2001.
30. Edworthy J, Loxley S, Dennis I. Improving auditory warning design: relationship between
warning sound parameters and perceived urgency. Hum Factors. 1991; 33:205-231.
468

31. Stanford L, McIntyre J, Hogan J. Audible alarm signals for anaesthesia monitoring
equipment. Int J Clin Monit Comput. 1985; 1:251-256.
32. Burt JL, Bartolome DS, Burdette DW, Comstock JR. A psychophysiological evaluation of
the perceived urgency of auditory warning signals. Ergonomics. 1995; 38:2327-2340.
33. Loeb RG. A measure of intraoperative attention to monitor displays. Anesth Analg. 1993;
76:337-341.
34. Gurushanthaiah K, Weinger MB, Englund CE. Visual display format affects the ability of
anesthesiologists to detect acute physiologic changes. A laboratory study employing a
clinical display simulator. Anesthesiology. 1995; 83:1184-1193.
35. Jungk A, Thull B, Hoeft A, Rau G. Ergonomic Evaluation of an Ecological Interface and a
Profilogram Display for Hemodynamic Monitoring. J Clin Monit. 1999; 15:469-479.
36. Kestin IG, Miller BR, Lockhart CH. Auditory alarms during anesthesia monitoring.
Anesthesiology. 1988; 69:106-109.
37. Sanderson P, Seagull FJ. Cognitive Ergonomics of Information Technology in Critical
Care: Contexts and Modalities for Alarm Interpretation. Available at:
http://www.detir.qld.gov.au/hs/ergo97/sanderso.pdf . Accessed June 20, 2001.
38. Xiao Y, Mackenzie F, Seagull J, Jaberi M. Managing the monitors: an analysis of alarm
silencing activities during an anesthetic procedure. In: Proceeding of the Human Factors
and Ergonomics Society 44th Annual Meeting. Santa Barbara: Human Factors and
Ergonomics Society, 2000: 250-253.
39. Bliss JP, Gilson RD, Deaton JE. Human probability matching behavior in response to
alarms of varying reliability. Ergonomics. 1995; 38(11):2300-2312.
40. Rheineck-Leyssius AT, Kalkman CJ. Influence of pulse oximeter lower alarm limit on the
incidence of hypoxaemia in the recovery room. Br J Anaesth. 1997; 79(4):460-464.
41. Makivirta A, Koski EM. Alarms-inducing variability in cardiac postoperative data and the
effects of prealarm delay. J Clin Monit. 1994; 10:153-162.
42. Rheineck-Leyssius AT, Kalkman CJ. Advanced pulse oximeter signal processing
technology compared to simple averaging. II. Effect on frequency of alarms in the
postanesthesia care unit. J Clin Anesth. 1999; 11:196-200.
43. Hyman WA. Errors in the Use of Medical Equipment. In: Bogner MS, editor. Human Error
in Medicine. Hillsdale, New Jersey: Lawrence Erlbaum Associates, 1994: 327-347.
44. Cohen T, Bakuzonis C, Friedman SB, Roa RL. Benchmark indicators for medical
equipment repair and maintenance. Biomed Instrum Technol. 1995; 29:308-321.
45. Cohen T. Validating medical equipment repair and maintenance metrics: a progress report.
Biomed Instrum Technol. 1997; 31:23-32.
46. Shaffer MJ. Biomedical equipment: maintenance factors, units of measurement, and
standards. Med Instrum. 1985; 19:268-272.
47. Leape LL. Error in medicine. JAMA. 1994; 272:1851-1857.
48. James PJ. Equipment management risk rating system based on engineering endpoints.
Biomed Instrum Technol. 1999; 33:115-120.
49. Cohen T. Validating medical equipment repair and maintenance metrics, Part II: Results of
the 1997 survey. Biomed Instrum Technol. 1998; 32:136-144.
50. Keil O, Widmann DE. Assessment of the impact of medical devices on the quality of care.
QRB Qual Rev Bull. 1984; 10:278-280.
51. Caplan RA, Vistica MF, Posner KL, Cheney FW. Adverse anesthetic outcomes arising
from gas delivery equipment: a closed claims analysis. Anesthesiology. 1997; 87:741-748.
469

52. Cooper JB, Newbower RS, Long CD, McPeek B. Preventable anesthesia mishaps: a study
of human factors. Anesthesiology. 1978; 49:399-406.
53. Cooper JB. Toward prevention of anesthetic mishaps. Int Anesthesiol Clin. 1984; 22:167-
183.
54. March MG, Crowley JJ. An evaluation of anesthesiologists' present checkout methods and
the validity of the FDA checklist. Anesthesiology. 1991; 75:724-729.
470

Chapter 42. Information Transfer
Harvey J, Murff, MD
David W. Bates, MD, MSc
Harvard Medical School
Introduction
Patient safety can be compromised by discontinuities in care. Studies suggest that
discontinuity results from poor information transfer 1 and faulty communication, 2 which in turn
may cause avoidable adverse events. 3
Improving information transfer and communications among health care providers is an
important patient safety practice and has been strongly recommended as a means to improve
patient care. 1,3-7 This chapter evaluates safety practices involving improvements of provider-toprovider
information transfer. Practices for evaluation include transfer of information between
inpatient and outpatient pharmacies (Subchapter 42.1), sign-out systems for medical housestaff
(Subchapter 42.2), automatically generated electronic discharge summaries (Subchapter 42.3),
and systems to improve patient notification of abnormal results (Subchapter 42.4).
Subchapter 42.1. Information Transfer Between Inpatient and Outpatient Pharmacies
Background
Accurate and timely information transfer between community and acute care pharmacies
is an important safety practice. Patients admitted to the hospital could benefit from the hospital’s
pharmacy obtaining better information concerning their medication allergies as well as prior
therapeutic failures. 8 Furthermore, when patients transition from acute care to outpatient care,
changes in medications that occurred during hospitalization may cause confusion for both
patients and providers. In one study surveying patients one week after hospital discharge,
patients’ knowledge of their drug indications were worse for medications introduced during their
hospitalization than for those taken prior to hospitalization (OR 0.69, 95% CI: 0.53-0.89). 9
Confusion and incomplete information may increase the risk of under- or
overmedication, harmful drug interactions, and other problems. Existing literature suggests that
pharmacist interventions may reduce potential adverse drug events and have a modest impact on
patient morbidity and mortality. 10,11 However, these studies have used independent reviewers to
judge the impact of the intervention and have not specifically measured adverse drug events or
patient outcomes. Clinical pharmacists’ consultations prior to discharge might also improve
patient medication compliance (see Chapter 7). 12
Uncontrolled studies report that information-exchange programs between hospital and
community pharmacies are perceived as beneficial and may have a positive impact on patient
outcomes. 13 Although not the primary outcome measured in their small (n=127) observational
study, Dvorak et al did note that using a pharmacy-to-pharmacy referral form was effective in
preventing 2 medication errors. Thus, practices that improve information transfer between
hospital and community pharmacies may improve patient safety.
Of the many potential methods for improving information transfer between hospitals and
outpatient pharmacies, controlled trials have been reported in the literature for only 2:
pharmaceutical care plans cards 14 and patient information facsimiles between pharmacies. 8
Although direct electronic communication of pharmacy data may be superior to these methods,
471

no controlled studies are currently available regarding this practice and therefore it is not
reviewed within this chapter.
Practice Description
In a study by Smith and colleagues, patients received a card prior to discharge listing
their pharmaceutical care plan, which included medication doses, indications, schedules, side
effects, information as to the importance of drug compliance, and how to obtain medication
refills. 14 Patients were instructed to give the card to their community pharmacist. In another
study, the intervention consisted of pharmacy-to-pharmacy facsimile transmission at the time of
admission and discharge from the hospital. 8 In this study, when a patient was admitted to the
hospital their community pharmacy transmitted patient demographic information, historical
information concerning allergies and adverse drug reactions, current medications, refill history,
pharmacist’s monitoring notes, communications with patient and physician, and a detailed
medication history to the admitting hospital. After discharge the hospital pharmacy transmitted
to the community pharmacy a list of any potential medication problems identified by the hospital
pharmacist on admission, the patient’s daily monitoring log, the pharmacist’s discharge
summary, and a discharge medication table. The medical records department of the hospital also
transmitted the patient’s discharge summary and laboratory test results to the community
pharmacy.
Prevalence and Severity of the Target Safety Problem
Medication problems can arise because patients are frequently discharged from the acute
care hospital on medications different from their ambulatory regimen. 15 Elderly patients in
particular are at risk after discharge. 16 In one study, hospital providers changed 53% of the drugs
prescribed by the primary care providers. 15
The extent to which these medication changes and lack of communication result in
recently discharged patients failing to receive medications or appropriate monitoring for their
drug therapy is unclear. In one study of elderly patients, 32% of medications prescribed at
discharge were not being taken 2 days after discharge. 17 Another study found that 51% of
patients recently discharged from acute care hospitals had deviated from their prescribed
regimen. 18 Of those that had deviated from the prescribed drug regimen, 70% did not understand
the medication regimen. In a Scottish study, recently discharged, elderly patients were issued a
5-day supply of their medication on discharge and visited in their homes after these 5 days had
elapsed. 16 Twenty-seven percent of the patients had not received a new prescription ordered on
discharge. Of the patients with new prescriptions issued, 19% received inaccurately labeled
medications. Medications were considered mis-labeled when non-specific container labels, such
as “take as directed” replaced the more specific labels given on discharge. Some authors have
suggested that improving communication about medications prior to and just after hospital
discharge might reduce these medication errors. 18,19
Poor communication is not the only problem. 16 Patient factors influence whether a
medication is ultimately picked up and taken as directed. Deviations from prescribed drug
regimens are multifactorial and improving pharmacy-to-pharmacy communication is only one
aspect of the overall problem.
Opportunities for Impact
Data from primary care providers reveal that 96% of the respondents would like
information concerning hospital drug changes. 20 Ninety-four percent of community pharmacists
472

surveyed also wished to be provided with information concerning hospital drug changes. 20 We
were unable to identify data regarding what percentage of hospital pharmacies routinely transfer
information on patients’ medication regimens when they are admitted to and discharged from
acute care.
Study Design and Outcomes
Two controlled studies were identified in the literature (Table 42.1.1). Both were
randomized trials but neither was blinded. In Smith et al, patients received a written pharmacy
care plan at discharge. Home visits were made 7 to 10 days later to assess compliance and
discrepancies in the medication that patients were taking versus those ordered at discharge
(Level 2). 14 In the study by Kuehl et al, 8 patients were randomly assigned to either usual care or
to a bi-directional exchange of pharmacy information by facsimile between the ambulatory
pharmacy and the admitting hospital, upon admission and discharge (Level 1). The outcomes
were pharmacist interventions, such as changing medication doses or making allergy
recommendations (Level 2). 8
Evidence for Effectiveness of the Practice
Smith et al’s small study (n=53) of a pharmacy care plan card found that in both groups
patients were taking different medications than those ordered at discharge. The authors found
that compliance with post-discharge medications was significantly better in the group that had
received the information card (p

group and the loss to follow-up group. This large loss to follow-up could have significantly
altered their findings.
Potential for Harm
None of the studies evaluating different hospital-community pharmacy communication
processes detailed any adverse events. However, the degree of added workload and effects on
current workflow must be taken into account.
Costs and Implementation
The 2 reviewed interventions for improving hospital to community pharmacy information
transfer seemed simple and relatively inexpensive. 8,8,13,14,14 Although no formal cost analysis was
performed, Dvorak et al described their method (use of a pharmacy-to-pharmacy referral form)
as being “labor-intensive.” No records were kept of the time necessary to provide the referrals
but the authors estimated the time commitment per patient to be 30 minutes. 13 It is also important
to note that complete information transfers occurred for 75% of the subjects in Kuehl’s study,
indicating that there is still room for improving the process. With improvements in information
transfer technology, automated transfer of information from hospital to community pharmacy
could have important patient safety benefits without excessively increasing providers’
workloads.
Comment
Providing patients with data forms to convey transfer of information between hospital
and ambulatory pharmacies has potential for reducing discontinuities resulting from inadequate
medication information. Few studies have evaluated the effect of these interventions on patient
outcomes, although any improvement in the transfer of this information would likely be well
received by ambulatory providers. Future studies are necessary to determine if and how
improved pharmacy-to-pharmacy communications reduce preventable adverse drug events and
improve patient outcomes. Identifying the most effective and least disruptive forms of interpharmacy
communication remains an area for further investigation.
474

Table 42.1.1. Practices to improve transfer of medication information
Study Study Setting Intervention Study Design,
Outcomes
Smith,
1997 14
Kuehl,
1998 8
References
53 patients (>65 yrs
old) discharged to
home who were
likely to experience
difficulties with their
medications
156 patients
admitted to small
mid-western
community hospital
Copies of medication
doses, indications, side
effects, importance of
compliance and refill
information given to
patients at discharge
Pharmacy-to-pharmacy
facsimile transmission o
medication regimen at
time of admission and
discharge from hospital
475
Level 1,
Level 2
Level 1,
Level 2
Results
Patients taking medication
not prescribed at discharge:
information card 75%, control
96% (p

13. Dvorak SR, McCoy RA, Voss GD. Continuity of care from acute to ambulatory care
setting. Am J Health Syst Pharm. 1998; 55:2500-2504.
14. Smith L, McGowan L, Moss-Barclay C, Wheater J, Knass D, Chrystyn H. An investigation
of hospital generated pharmaceutical care when patients are discharged home from
hospital. Br J Clin Pharmacol. 1997; 44:163-165.
15. Himmel W, Tabache M, Kochen MM. What happens to long-term medication when
general practice patients are referred to hospital? Eur J Clin Pharmacol. 1996; 50:253-257.
16. Burns JM, Sneddon I, Lovell M, McLean A, Martin BJ. Elderly patients and their
medication: a post-discharge follow-up study. Age Ageing. 1992; 21:178-181.
17. Beers M, Sliwkowski J, Brooks J. Compliance with medication orders among elderly after
hospital discharge. Hosp Formul. 1992; 27:720-724.
18. Parkin D, Henney C, Quirk J, Crooks J. Deviations from prescribed drug treatment after
discharge from the hospital. BMJ. 1976; 2:686-688.
19. Katz E, Nicod P, Brunner HR, Waeber B. Changes in treatment during and after
hospitalization in patients taking drugs for cardiovascular disease. Cardiovascular Drugs
and Therapy. 1996; 10:189-192.
20. Munday A, Kelly B, Forrester JW, Timoney A, McGovern E. Do general practitioners and
community pharmacists want information on the reasons for drug therapy changes
implemented by secondary care? Br J Gen Pract. 1997; 47:563-566.
Subchapter 42.2. Sign-Out Systems for Cross-Coverage
Background
As physicians go off duty, they provide information to a “cross-covering” physician who
will care for patients in the interim. The process of information transfer, known as “sign-out,” is
often informal and unstructured. Various methods are used, including handwritten lists, PCbased
word processing or spreadsheet programs, and personal digital assistants (PDAs), but little
literature has assessed their effectiveness in assuring continuity of care for patients and
preventing medical errors. Although notes in the medical record often contain all the information
needed to care for patients, cross-covering physicians make many decisions without the benefit
of the patients’ charts. 1 Jelley found lack of consistency in the content of weekend sign-out lists
in a community-based internal medicine inpatient program. 1 Lee and colleagues found that
medical interns recorded information elements such as patient age, DNR status, and medications
more often when a standardized sign-out card was used. 2 In this section, we review evidence of a
computerized sign-out system to reduce medical errors during cross-coverage.
Practice Description
The proposed safety practice is a structured sign-out process in which patient information
is provided for various standardized data fields. The computerized sign-out program described
by Peterson and colleagues consisted of a summary of the patient’s medical status, a problem
list, recent laboratory data, resuscitation status, allergies, and a “to do” list. 3 This information
was accessible from any computer within the hospital and was accessed and maintained on a
daily basis by housestaff physicians.
Prevalence and Severity of the Target Safety Problem
Discontinuities in provider care during hospitalization have been associated with an
increased risk of adverse events. The Petersen et al study found the odds ratio for a preventable
476

adverse medical event occurring during cross-coverage as opposed to regular provider coverage
to be 6.1 (95% CI: 1.4-26.7). 4 In the surgical domain, the Joint Commission on Accreditation of
Healthcare Organizations (JCAHO) publication on wrong-site surgeries noted a number of cases
involving last minute personnel changes. 5 It is possible that these rapid substitutions in the
operating room with inadequate communication may have contributed to these adverse events.
Opportunities for Impact
The number of hospitals using either a computerized or paper-based sign-out process is
unknown, but computerized sign-outs are probably unusual. One study found that 26% of
adverse events in a single institution occurred during cross-coverage. 4 These data suggest that a
standardized sign-out procedure could have a significant impact on improving patient safety.
Study Design and Outcomes
We identified one study evaluating the effect of a standardized sign-out system on the
occurrence of adverse events (Table 42.2.1). In this study, adverse medical events detected by
self-report were compared before and after implementation of a computerized sign-out system
(Level 3). 3 This study involved internal medicine housestaff and any management of a patient
performed by an intern from a different team or a night-float resident was considered crosscoverage.
During chart review, the investigators recorded whether the physician at the time of
the event was the patient’s regular physician or a cross-covering physician. Adverse medical
events, defined as “an injury due to medical therapy that prolonged hospital stay or disability at
discharge” 3 were the primary outcomes (Level 2).
Evidence for Effectiveness of the Practice
There were significantly fewer adverse events during the intervention period compared
with the baseline period (2.38% vs. 3.94%, p

Potential for Harm
The study reported no adverse events as a result of the sign-out system. As with other
sign-out systems, particularly those that are computerized or Web-enabled, the issues of data
security and protection of confidentiality must be addressed. 6
Costs and Implementation
Implementation of a computerized sign-out system like that described by Peterson et al
would require information systems that allow extraction and aggregation of patient specific data
(eg, laboratory and pharmacy) and financial support for programming. These resources may not
be present at some institutions. Although housestaff responded favorably to the computerized
system, physicians at other institutions or in other specialties may not be as willing to use this
system.
Comment
One study has shown that an inpatient’s risk of preventable adverse events was less after
implementation of a computerized sign-out process. The method appears appropriately suited for
hospitals with cross-coverage arrangements similar to those described by Peterson et al,
specifically in-house coverage by resident trainees. It will be important to know if similar
systems are as effective and well received at other institutions, including those without trainees.
Such systems would be difficult to implement in hospitals with limited information systems or
where physicians outside the hospital provide coverage through paging systems. Although a
computerized system has the advantage of being accessible from any location in the hospital and
may be able to automatically import important information, events attributable to faulty
communication during cross-coverage could also be amenable to other strategies for
standardizing the process (eg, sign-out cards, Web-based programs, PDAs). No evidence is
available concerning the relative effectiveness of other standardized sign-out methods. Future
research should address what data fields are most helpful to physicians providing cross-coverage
in preventing adverse events and how different methods of standardized sign-out compare in
effectiveness (eg, handwritten cards vs. PDAs).
478

Table 42.2.1. Computerized sign-out program
Study Setting Study Design,
Outcome
8767 patients admitted to
the medical service of a
tertiary care teaching
hospital in Boston 3
Level 3,
Level 2
Results (95% Confidence Intervals)
Odds ratio (OR) of preventable adverse events
occurring during cross-coverage compared with
care under regular physician:
baseline, OR 5.2 (1.5-18.2)
with intervention, OR 1.5 (0.2-9.0)
References
1. Jelley MJ. Tools of continuity: the content of inpatient check-out lists. J Gen Intern Med.
1994;9:77.
2. Lee LH, Levine JA, Schultz HJ. Utility of a standardized sign-out card for new medical
interns. J Gen Intern Med. 1996;11:753-755.
3. Petersen LA, Orav EJ, Teich JM, O'Neil AC, Brennan TA. Using a computerized sign-out
program to improve continuity of inpatient care and prevent adverse events. Jt Comm J
Qual Improv. 1998;24:77-87.
4. Petersen LA, Brennan TA, O'Neil AC, Cook EF, Lee TH. Does housestaff discontinuity of
care increase the risk for preventable adverse events? Ann Intern Med. 1994;121:866-872.
5. The Joint Commission on Accreditation of Healthcare Organizations. Sentinel event alert.
Lessons learned: wrong site surgery. 1998. Available at:
http://www.jcaho.org/edu_pub/sealert/sea6.html. Accessed June 27, 2001.
6. Balsbaugh TA. The family physician and handheld computers: a beginner's guide. The
Internet Journal of Family Practice 1[2]. 2001.
Subchapter 42.3. Discharge Summaries and Follow-up
Background
Discharge summaries are important tools for communicating pertinent patient
information regarding hospitalizations to outpatient care providers. Yet their relatively
unstructured, narrative format often invites inaccuracies. 1 In addition, there can be significant
delays transmitting discharge summaries to patients’ health care providers. 2,3
Prior studies have investigated processes to improve discharge summaries, such as
standardizing their format 4-6 and instituting physician education programs. 7 This chapter focuses
on the use of structured, database-generated discharge summaries to improve the quality of the
information content communicated after patient discharge, as well as to reduce the time required
for this information transfer. 8
Practice description
During the hospital course, physicians provide information corresponding to specific
sections of the computerized discharge summary either on data collection forms which are
manually entered into a database 8 or directly into a computer system. When the patient is
discharged the database generates a structured discharge summary that can be sent to the
patient’s outpatient providers.
479

Prevalence and Severity of the Target Safety Problem
In one study examining the effectiveness of inpatient follow-up care, 9.7% of discharged
patients experienced worsening of symptoms or functional capacity as a result of an inadequately
managed discharge process. 2 Hospital discharge summaries are an important means of
communication between hospital and community physicians, but have several problems. First,
community physicians do not always receive summaries for recently discharged patients. In one
study only 34% of patients had a discharge summary sent to their outpatient care provider. 2
Although no analysis was undertaken to determine if receiving a discharge summary had an
effect on patients’ follow-up, another study demonstrated that patients may be less likely to be
readmitted to the hospital if their primary care provider receives a discharge summary. 9
As mentioned above (Subchapter 42.1), patients frequently have their medication
regimen changed while admitted. 10 The majority of ambulatory providers would like to have
information regarding these medication changes. 11 Improvement in information transfer from
acute care to ambulatory care might reduce medication discrepancies; however, patient
compliance will also heavily influence these factors.
Opportunities for Impact
We found no data describing how many hospitals currently use database-driven discharge
summaries.
Study Design and Outcomes
Several studies were identified that evaluated electronically generated discharge
summaries, but these were limited by a lack of randomization or limited outcomes reporting. 12-14
One randomized controlled trial was identified that compared traditional dictated discharge
summaries to summaries generated from a database (Table 42.3.1). 8 The primary outcome was
the proportion of admissions with a discharge summary completed by 4 weeks after discharge
(Level 3). 8 Overall quality of the discharge summaries was also assessed (Level 3) but patient
level outcomes were not.
Evidence for Effectiveness of the Practice
Patients randomized to the database group were significantly more likely to have a
discharge summary generated within 4 weeks of discharge than were patients randomized to the
dictation group (113/142 vs. 86/151, p

Costs and Implementation
The direct and indirect costs of implementing and maintaining a system for databasegenerated
discharge summaries have not been formally evaluated in the literature. Results of a
mail survey of housestaff in the van Walraven study 8 suggest their intervention did not adversely
affect the workflow of housestaff physicians. Housestaff significantly preferred (p

6. Rawal J, Barnett P, Lloyd BW. Use of structured letters to improve communication
between hospital doctors and general practitioners. BMJ 1993; 307:1044.
7. Flyer B, Rubenstein LZ, Robbins AS, Wieland GD, Henry D, Cugalj N. An intervention to
improve the hospital discharge summary. J Med Educ 1988; 63:407-409.
8. van Walraven C, Laupacis A, Seth R, Wells G. Dictated versus database-generated
discharge summaries: a randomized clinical trial. CMAJ 1999; 160:319-326.
9. van Walraven C, Seth R, Austin P, Laupacis A. The effect of discharge summaries on
readmission to hospital. 2001. [unpublished work]
10. Himmel W, Tabache M, Kochen MM. What happens to long-term medication when
general practice patients are referred to hospital? Eur J Clin Pharmacol 1996; 50:253-257.
11. Munday A, Kelly B, Forrester JW, Timoney A, McGovern E. Do general practitioners and
community pharmacists want information on the reasons for drug therapy changes
implemented by secondary care? Br J Gen Pract 1997; 47:563-566.
12. Lissauer T, Paterson CM, Simons A, Beard RW. Evaluation of computer generated
neonatal discharge summaries. Arch Dis Child 1991; 66(4 (Spec No)):433-436.
13. Llewelyn DE, Ewins DL, Horn J, Evans TG, McGregor AM. Computerised updating of
clinical summaries: new opportunities for clinical practice and research? BMJ 1988;
297:1504-1506.
14. Smith RP, Holzman GB. The application of a computer data base system to the generation
of hospital discharge summaries. Obstet Gynecol 1989; 73(5 Pt 1):803-807.
15. Maresh M, Beard RW, Combe D, Dawson AM, Gillmer MDG, Smith G et al. Selection of
an obstetric data base for a microcomputer and its use for on-line production of birth
notification forms, discharge summaries and perinatal audit. Br J Obstet and Gyn 1983;
90:227-231.
Subchapter 42.4. Notifying Patients of Abnormal Results
Background
One of the most distressing safety issues of the clinical encounter is the failure to followup
on diagnostic tests, particularly when a patient is not notified of an abnormal result. The
complexities of this problem are legion. Contact methods—whether by phone, mail, fax or email,
and whether sent by the lab, clinic or individual clinician—vary widely in their reliability
(with most being imperfect). In some instances patients are told that if they do not hear back
regarding their test results it signifies normal results. Of course, not hearing may mean that the
test was lost or that the contact method was faulty. Other issues arise when the content of the
notification is not clear, either as to result or the recommended follow-up for re-testing or
treatment options.
This chapter evaluates safety practices aimed at improving patient notification of
abnormal results. Adequate medical and/or surgical care during follow-up is essential to
reducing patient morbidity and mortality, but practices to address this are beyond the scope of
patient safety as defined in this Report. We have chosen the example of Pap smear results,
although many of the issues should be transferable to other laboratory (eg, PSA level) and
radiologic (eg, mammogram) results.
482

Practice Description
Our search revealed only one study evaluating patient notification practices. 1 In this
study, the patient’s mailing address was included on the Pap smear request form. Two weeks
after the patient’s primary care provider received the results, the laboratory directly notified the
patient by mail. The notification was by form letter advising the patient of her results and
providing advice on the recommended follow-up step: discuss results with the doctor, return in
two years time, or make an appointment to see the doctor without delay.
Prevalence and Severity of the Target Safety Problem
Few data exist concerning physician follow-up and patient notification of abnormal
results. In a survey of attending physicians and residents practicing at a large urban teaching
hospital and 21 suburban primary care practices, virtually all respondents believed it was
moderately or extremely important to notify patients of abnormal results, yet 36% of physicians
did not always do so. 2 Among the most common reasons reported by physicians were
forgetfulness and inability to reach patients. One large cross-sectional study examined physician
documentation of notification to patients of abnormal mammograms, Pap smears, and
cholesterol tests. The results demonstrated that certain patient characteristics such as race,
language, and education may be associated with a failure to transmit abnormal results to
patients. 3
An estimated 4600 American women died of cervical cancer in 2000. 4 There are no data
regarding to what extent delays in notification result in worse patient outcomes, including
mortality. One study evaluating processes to reduce non-adherence rates with the follow-up of
abnormal Pap smears noted that many women (the exact number was not presented) reported
that they were never notified of their abnormal Pap smear result initially. 5
Tracking systems for abnormal Pap smear results have been briefly mentioned in the
context of studies evaluating interventions to improve overall follow-up, not patient
notification. 5,6 However, the effectiveness of these tracking systems was not specifically
evaluated so they are not reviewed here.
Opportunities for Impact
Compliance rates with follow-up medical care after abnormal Pap smears typically
ranges from 50% to 70%. 5,7-9 It is unclear how often losses to follow-up resulted from failure to
notify. There are no data indicating how many practice groups directly mail abnormal Pap smear
results to patients. Even with successful patient notification practices, a corresponding reduction
in morbidity and mortality may not occur because of the other barriers to adequate follow-up
described in the literature. 5,10
Study Design and Outcomes
The study reviewed for this chapter was a randomized control design (Level 1). 1 (Table
42.4.1). Providers were randomized into 2 groups. In the intervention group, the pre-cervical
smear questionnaire form had been redesigned to allow the patient to request that results be
mailed directly to her. The physicians in the intervention group determined which patients would
be offered direct notification. Patients of physicians in the control group were notified of results
using whatever protocol the provider typically used. The authors did not elaborate on the
methods used to notify patients in the control group.
483

The primary outcome was adherence with follow-up visits (Level 2), which was defined
by searching the laboratory records for evidence of a follow-up Pap smear one year after
notification. If a cervical smear result was not located through the laboratory database search
then the patient’s provider was contacted. 1
Evidence for Effectiveness of the Practice
Significantly fewer women with cervical intraepithelial neoplasia (CIN) on Pap smear
who were randomized to the intervention group were lost to follow-up (0/52 vs. 9/39 in the
control group, p

Table 42.4.1. Randomized controlled trial of direct notification of abnormal Pap smear
results*
Study Study Setting Outcomes Results (95% Confidence Interval)†
Del Mar,
1995 1
311 women with abnormal
Pap smears from 42
general practices in
Australia
Level 2 Patients with CIN lost to follow-up:
Direct mail notification: 0 (0-0.07)
Control: 0.23 (0.11-0.39)
Patients with atypia lost to follow-up
Direct mail notification: 0.10
Control: 0.13 (p=NS)
* CIN indicates cervical intraepithelial neoplasia; NS, not statistically significant.
† Proportion of patients lost to follow-up reported in intervention group vs. control group.
References
1. Del Mar CB, Wright RG. Notifying women of the results of their cervical smear tests by
mail: does it result in a decreased loss to follow-up of abnormal smears? Aust J Public
Health 1995; 19:211-213.
2. Bookhaker EA, Ward RE, Uman JE, McCarthy BD. Patient notification and follow-up of
abnormal test results: a physician survey. Arch Intern Med 1996; 156:327-331.
3. Burstin HR, Puoplolo AL, Haas JS, Cook EF, Brennan TA. Patients at risk for missed
follow-up after abnormal tests [abstract]. J Gen Intern Med 13, 46A. 1998.
4. American Cancer Society. Cancer Facts & Figures 2000. Atlanta, GA . 2000.
5. Marcus AC, Kaplan CP, Crane LA, Berek JS, Bernstein G, Gunning JE et al. Reducing
loss-to-follow-up among women with abnormal Pap smears. Results from a randomized
trial testing an intensive follow-up protocol and economic incentives. Med Care 1998;
36:397-410.
6. Paskett ED, White E, Carter WB, Chu J. Improving follow-up after an abnormal Pap
smear: a randomized controlled trial. Prev Med 1990; 19:630-641.
7. Marcus AC, Crane LA, Kaplan CP, Reading AE, Savage E, Gunning J et al. Improving
adherence to screening follow-up among women with abnormal Pap smears: results from a
large clinic-based trial of three intervention strategies. Med Care 1992; 30:216-230.
8. Michielutte R, Diseker RA, Young LD, May WJ. Noncompliance in screening follow-up
among family planning clinic patients with cervical dysplasia. Prev Med 1985; 14:248-258.
9. Singer A. The abnormal cervical smear. Br Med J (Clin Res Ed) 1986; 293:1551-1556.
10. Stewart DE, Buchegger PM, Lickrish GM, Sierra S. The effect of educational brochures on
follow-up compliance in women with abnormal Papanicolaou smears. Obstet Gynecol
1994; 83:583-585.
11. Stewart DE, Lickrish GM, Sierra S, Parkin H. The effect of educational brochures on
knowledge and emotional distress in women with abnormal Papanicolaou smears. Obstet
Gynecol 1993; 81:280-282.
12. Paskett ED, Phillips KC, Miller ME. Improving compliance among women with abnormal
Papanicolaou smears. Obstet Gynecol 1995; 86:353-359.
485

Final Comment to Chapter 42
Faulty information transfer causes discontinuities of care that may result in adverse
events. However, interventions to improve information transfer have received relatively little
attention in the medical literature. Unfortunately, numerous barriers impede the appropriate
transfer of information between institutions and between patient and provider. Future
technologies that allow for more seamless transfer of information may mitigate these gaps in
patient care. Further evaluation is critical.
486

Chapter 43. Prevention of Misidentifications
Heidi Wald, MD
University of Pennsylvania School of Medicine
Kaveh G. Shojania, MD
University of California, San Francisco School of Medicine
Subchapter 43.1. Bar Coding
Background
Machine-readable automatic identification (ID) systems, including bar codes, magnetic
stripes, optical character recognition and radiofrequency labeling, have improved productivity
and quality in diverse industries. Bar codes represent the oldest and most common of these
machine-readable ID systems, 1,2 and are widely used in industrial manufacturing, shipping and
inventory tracking operations. Prior to bar coding, these processes would have involved
keystroke entry of identification numbers, producing approximately one error in every 300
entered characters. In contrast, bar coding produces misidentification errors at rates ranging from
one character in 15,000 to one character in 36 trillion. 3
The use of bar coding in health care was first described over 30 years ago in clinical
laboratories and blood banks. 1,4 In 1984, Rappoport identified 3 areas for the use of automatic ID
technology in lab medicine: patient identification, document identification, and specimen
identification. 1 However, in a 1987 survey by the American Hospital Association, the use of bar
codes was most widespread in materials management departments, rather than in clinical
application. 5 Other areas in which hospitals employed bar codes at that time included the clinical
laboratory, pharmacy, radiology, medical records and asset management. Despite the Health
Industry Bar Code Council’s call for standardization in the mid-1980s, the implementation of bar
code technology has been stymied by lack of industry standards and failed cooperation among all
stakeholders. 2,6
As bar coding has the potential to substantially increase productivity and accuracy, one
would expect it to be applied to important patient safety practices. Unfortunately, the published
literature contains very little evidence regarding health care applications. In this chapter we focus
on 4 areas in which bar coding shows promise for improving patient safety: patient
identification, medication dispensing and administration, specimen handling, and medical record
keeping. 2,7-10
Practice Description
Patient Identification
Machine-readable patient ID systems could replace conventional wrist-banding, and
might reduce patient ID errors ranging from specimen collection to medication and blood
product administration. Attempts to create such machine-readable ID systems were reported in
the blood banking literature as early as 1977. 11 Transfusion Medicine is particularly attuned to
issues of patient identification, as specimen collection and blood product administration account
for the majority of preventable transfusion errors. 12,13
Most scenarios for patient identification involve the substitution or supplementation of
the traditional wristband for one with a unique bar code patient identifier. All patient specimens,
medications and released blood products then receive the patient’s unique bar code ID. No
487

procedure or treatment can occur unless the patient’s ID is scanned with a portable scanner and
matched with a bar code generated by the doctor’s order. For example, a phlebotomist would
carry the scanner, check the patient’s ID against a bar coded specimen label or collection list,
and draw blood only in the event of a match. Similarly, for administration or treatment, the
patient’s ID and the intended therapeutic would be scanned at the bedside with a portable reader.
If a match exists, the transfusion or medication is allowed and the time and date are recorded and
even transmitted directly to the hospital computer system. The nurse’s bar code ID can also be
scanned and a timed administration record can be created. If there is no match, an alarm is
sounded, and the administration delayed until the problem is resolved. 9,14
Other technological means to reduce error in patient identification have been examined in
the transfusion literature. Several researchers explored the use of a system providing a
mechanical barrier to transfusion through a series of locks. 15 This system appears to be
cumbersome and easily circumvented. In addition, electronic blood banking, with point-of-care
crossmatching and computer-controlled release of blood, has been examined for high volume
transfusion areas such as the operating room or intensive care unit. 16 Due to major barriers to
large-scale implementation, these practices will not be discussed further.
Specimen handling
Clinical laboratories have integrated bar codes in specimen handling with a great deal of
success. 3,8 Several authors have described the development of central laboratory information
systems (LIS) that employ bar code technology. Collection list and label software can be
modified to integrate and produce bar coded information. Confirmation of labeling and patient
ID occurs at the bedside. Specimen sorting and aliquoting in the laboratory can be shortened or
eliminated by various setups. At Rush-Presbyterian Hospital in Chicago, a central receiving
station rapidly sorts and aliquots bar coded samples as they move on a conveyor belt. 3 At the
University of Kansas Medical Center, the collection tubes are also used for analysis, thus
eliminating the need for aliquoting samples. Additionally, the computer sends the bar codecoordinated
orders to each of the 2 chemistry analyzers. Because a sample can then be
appropriately processed at either analyzer, the need for sorting samples has also been
eliminated. 8 Clinical labs also employ bar code technology in harder-to-automate processes. For
instance, the University of Utah uses bar codes to replace common keystrokes for text reporting
of microbiology results—eg, a technician might use a bar code “pick list” to scan the single bar
code that means “no growth for 24 hours” and eliminate the need to type this phrase. 17,18
Medication dispensing and administration
The use of bar codes is uniquely suited to the dispensing and administration stages of the
medication process. 19 Bar coding may be used to simplify the patient cassette (the medicine tray
for bedside delivery) filling and verification process. 20 For instance, a technician fills a cassette
according to a computerized, bar coded medication schedule and completes a quick verification
by scanning the label of each unit-dose that has been placed in it with a handheld scanner. The
computer can generate an error message if an incorrect medication is entered. The administration
of bar coded medications can also be tracked at the point-of-care using a portable scanner and
compared against the hospital computer’s medication orders. 9,21 Bar coded medication
administration has the added capability of creating a record of the administration (ie, RN, date,
time) and a bill. This type of system could be integrated with a patient identification system in an
attempt to eliminate errors resulting in administration of medication to the wrong patient.
488

Medical record keeping
Radiology and medical records departments use bar code technology to track the location
and status of studies and charts. 22-24 Even more creative use of bar coded information has been
reported in the emergency medicine and pharmacy literatures. As with the applications in the
microbiology lab, bar codes can be used to replace frequently used text for the creation of
medical records. “Pick lists” of bar codes with their text equivalents can be employed in
circumstances requiring speed and accuracy. Several uses of bar coded scripts have been
examined in resuscitation events, and in mass casualty recording. 10,25-27 The use of bar code
“pick lists” for the documentation of pharmacists’ clinical activities has also been explored. 28-30
Prevalence and Severity of the Target Safety Problem
Bar code technology may be used to address any number of patient safety issues in
medicine. For this discussion, we will define the target safety problem as patient identification in
general, using transfusion medicine as a specific example.
Patient identification remains a challenge in hospitals because of the number of complex
interventions that occur to patients ranging from meals to surgeries. These interventions occur in
a variety of locations and are provided by large teams of staff who work in shifts. In addition,
sick patients, or those who have a language barrier, are not always capable of responding to
questions about their identity or treatment plans. Hospitals generally rely on standardized
wristbands containing the patient’s name and other identifying information such as medical
record number or date of birth. Unfortunately, conventional wristbands are not reliable sources
of patient identification. A 1991 national sample of 712 hospitals estimated error rates for
conventional patient identification wristbands to be 5.5%. 31 In half of the errors, the patient’s
wristband was absent altogether. The error rates were significantly lower in hospitals where
phlebotomists had responsibility for monitoring wristband accuracy, as the phlebotomy staff
would not perform routine lab work unless the band was corrected. Other errors included more
than one wristband with conflicting data (18.3%); wristbands with incomplete (17.5%),
erroneous (8.6%), or illegible data (5.7%); and rarely, patients wearing wristbands with another
patient’s data (0.5%). As patient identification data are only as good as the information entered
at registration, the use of bar coded ID data could not be expected to correct certain types of
errors such as a wristband with incorrect data entered at admission, although it is potentially
beneficial in eliminating other types of errors such as illegible data.
Even when wristbands are free of errors, protocols for patient identification (such as dual
witness verification of identification for blood transfusion) are easily circumvented or performed
incorrectly. 32 In an analysis of major transfusion errors reported to the FDA over a 10-year
period from 1976-1985, Sazama found 10 patient deaths where the actual and intended patients
shared the same last name, and 5 deaths where the 2 shared the same hospital room. 13
Consequently, automatic patient identification systems have been proposed as a technological
solution to remove human factors (Subchapter 41.1) from the patient identification process.
Despite technical improvements in testing for blood group identification, fatal ABOincompatible
transfusions in the United States continue to occur at a rate ranging from
approximately 1:600,000 to 1:800,000, with as many as two dozen fatalities in the US
annually. 12,13 Thus, the chance of a patient suffering a fatal transfusion reaction due to ABOincompatibility
is roughly equivalent to the risk of acquiring HIV infection from a blood
transfusion. 33,34 Patient misidentification represents the most common cause of ABOincompatible
transfusion, accounting for 46-57% of these errors. 12,13 Since the rate of patient and
489

donor having blood group compatibility by chance is approximately 60%, it is estimated that the
total number of ABO-incompatible transfusions is much higher than the rate of fatal errors. A
study from New York State estimated that as many as one in 12,000 transfusions involve
administration of a blood product intended for another patient or release of blood of an incorrect
group. *12
Opportunities for Impact
The implementation of automatic patient identification may present a large opportunity
to bring transfusion medicine and other hospital interventions closer to the goal of zero risk.
According to a survey conducted by the American Society of Hospital-System Pharmacists,
1.1% of responding hospitals use bar coding of drug products in conjunction with bar coding on
the patient’s identification tag. 35
Study Designs
Multiple reports of the use of bar codes appear in the medical literature, but most of these
relate to inventory management. Few authors examine bar codes for patient identification. Only
one of these studies was a prospective evaluation, and in the study bar coded patient
identification comprised only one small part of the intervention. 10 One observational study
examined a bar code patient ID system for medication administration. 9 In the study, however,
routine patient ID scanning was easily circumvented and the actual error rate was not provided.
The remainder of the reports are descriptive in nature. 14,36,37 Therefore, error rates in automated
patient identification could not be readily compared to usual practice.
The use of bar coding in other clinical care applications (aside from inventory control)
has been examined prospectively in trauma recording 27 and in documenting pharmacists’
interventions. 28 One additional observational study examined bar coding in pharmacy
dispensing. 20
*This is based on a reported error rate of 1/19,000 transfusions. The authors estimate a 64%
chance that a random transfusion to an unintended recipient would be compatible and assume
100% reporting of incompatible erroneous transfusions
490

Study Outcomes
Some of the studies report error rates in transfusion or medication errors (Level 2), while
others report related outcomes such as speed of data entry (Level 3), transcription errors (Level
2) in a variety of experimental and clinical settings, and user satisfaction (Level 3).
Evidence for Effectiveness of the Practice
A point-of-care information system for medication management was implemented at a
tertiary care center in Colorado. 9 The system provided online patient and medication data that
verified medication administration at bedside using hand-held scanners to record patient ID,
nurse ID, and medication unit-dose bar codes. When intervention data were compared with
historical controls, the pre-intervention medication error rate of 0.17% dropped to 0.05%,
sustained over 3 years for an overall decrease of 71% (p value not reported). There was a 33%
decrease in "wrong drug" errors, a 43% decrease in "wrong time" errors, and a 52% decrease in
"omitted dose" errors. There was a 47% decrease in "transcription/order-entry" errors. There was
no change in "wrong patient" errors or "wrong dosage" errors, perhaps because the component of
the multifaceted intervention most likely to mitigate these errors, the use of the scanners for
patient ID, was easily and frequently circumvented. It is unclear if bedside scanning added an
unwanted layer of work for the nurses, or if they were uncomfortable performing this task in
front of patients and families. Computerized pharmacy and bar code tracking of medications led
to qualitative improvements in documentation time, scheduling of administration, nursingpharmacy
communications, and pharmacist drug monitoring. However, the contribution of bar
coding to the decreased error rate is not distinguishable from that of the entire intervention. It
also appears that the point-of-care patient ID portion of the intervention was easily bypassed, and
was therefore not adequately evaluated. 9
On a large medical ward of a university hospital, the satellite pharmacy was reconfigured
to implement bar code technology for drug dispensing. 20 The hospital bar coded all medications,
patient medication cassettes, patient wristbands, and employees. Standard dispensing time was
estimated at 8.24 seconds, while dispensing time for bar coded medications was 6.72 seconds (p
value not reported). Accuracy of the standard cassette fill system was 99.6% (equivalent to one
error in 250 doses), while the accuracy of bar coded cassette fill was reported to be 100% (based
on 0 errors in about 20,000 doses). The pharmacists were freed to do other work as the burden of
dispensing was shifted to pharmacy technicians.
As pharmacists have begun to use bar code technology for medication distribution, 2
pharmacy groups described the use of bar codes in recording their clinical interventions. 28, 29 One
group found bar code documentation to have lower overall error rates compared to manual
documentation, while the marginal cost of implementing bar coding was less than $100
(factoring in savings in labor costs related to manual entry). 28 These data were limited by the
small number of operators who differed on their preferences for manual versus bar code
recording.
A prospective trial reviewed bar code technology in trauma recording. Experienced
emergency room nurses found bar coded pick lists to be easy to use and produced fewer errors
per record compared with handwriting (2.63±0.24 vs. 4.48±0.3, p

The limitations of these studies are numerous, including their small sample sizes and lack
of generalizability. However they demonstrate that bar code technology is generally easy for
operators to use, can be applied in a variety of creative ways, and produces easy-to-demonstrate
gains in accuracy and efficiency.
Potential for Harm
There is no clear detriment to patient identification with a bar code system. However, as
with the addition of any new technology, the possibility exists that the complexity of the
information system, especially if it grows, could create more routes for potential failure. For
instance, an error during patient registration might be perpetuated throughout the hospitalization
by the information systems, and be more difficult to correct than with conventional systems. The
system’s data are only as accurate as that entered by fallible humans.
Costs and Implementation
Significant barriers need to be overcome before the full potential of bar coding can be
exploited in the clinical setting. The process of medication dispensing and administration
highlight several examples of these barriers. First, pharmaceutical manufacturers have yet to
adopt a universal bar code standard like the UPC system used in grocery store inventory. 38 As of
yet, there has been no regulatory mandate by the FDA to serve as an incentive, although the
American Society of Hospital-System Pharmacists recently urged the FDA to take action. 39
Second, placing bar codes on unit-doses of medications often requires major changes in
packaging (such as an increase of the package size to accommodate the bar coded label). Bar
code tracking systems would have difficulty with unusual doses such as halved tablets. IV doses
would still require bar code labeling in the pharmacy when prepared.
At this point, implementation of bar coding requires a commitment on the part of the
hospital to relabel every unit-dose of medication using a hospital standard. The hospital
pharmacies that have implemented these systems have repackaged and relabeled many unitdoses
at considerable cost. 9,20 One health system estimated costs at $119,516 annually, with the
per dose costs of bar code labeling estimated at 2.73 cents. 20 The bottom line is that, at present,
the costs of implementing bar coding in an entire pharmacy inventory are significant and the
logistics complex.
The use of bar coding in simpler clinical scenarios (ie, using a blood transfusion
wristband for patient identification) may be implemented with very modest outlay of resources
(estimated to be less than 5 cents per wristband). 36 The costs of the scanners themselves are
moderate. A new model scanner, software and recharger were priced at about $1100 in 1997. 28
Comment
Bar coding is a fast and accurate method of automated data capture that in experimental
settings provides qualitative improvements in speed and accuracy of data entry. Bar code
technology might be creatively applied to any number of data-driven processes in medicine. As
the rapid and accurate transfer of data is paramount in health care, the thoughtful application of
appropriately piloted and evaluated bar code technology is likely to be well received and
deserves further investigation.
492

References
1. Rappoport A. A hospital patient and laboratory machine-readable identification system
(MRIS) revisited. J Med Syst. 1984;8:133-156.
2. Weilert M, Tilzer LL. Putting bar codes to work for improved patient care. Clin Lab Med.
1991;11:227-238.
3. Maffetone MA, Watt SW, Whisler KE. Automated sample handling in a clinical
laboratory. Comput Healthc. 1988;9:48-50.
4. Brodheim E, Ying W, Hirsch R. An evaluation of the codabar symbol in blood bank
automation. Vox Sang. 1981;40: 3.
5. Longe K. The status of bar codes in hospitals: a survey report. Hospital Technology Series.
Chicago: American Hospital Association; 1989. 8.
6. Garza D, Murdock S, Garcia L, Trujillo JM. Bar codes in the clinical laboratory. Clin Lab
Sci. 1991;4:23-25.
7. Zarowitz BJ, Petitta A, Rudis MI, Horst HM, Hyzy R. Bar code documentation of
pharmacotherapy services in intensive care units. Pharmacotherapy. 1996;16:261-266.
8. Tilzer LL, Jones RW. Use of bar code labels on collection tubes for specimen management
in the clinical laboratory. Arch Pathol Lab Med. 1988;112:1200-1202.
9. Puckett F. Medication-management component of a point-of-care information system. Am
J Health-Syst Pharm. 1995;52:1305-1309.
10. Noordergraff G, Bouman J, Van Den Brink E, Van De Pompe C, Savelkoul T.
Development of computer-assisted patient control for use in the hospital setting during
mass casualty incidents. Am J Emerg Med. 1996;14:257-261.
11. Sherer P, Chambers R, Taswell H, Altshuler C, Aster R, Covino K, et al. Automated donorrecipient
identification systems as a means of reducing human error in blood transfusion.
Transfusion. 1977;17:586-597.
12. Linden J, Paul B, Dressler K. A report of 104 transfusion errors in New York State.
Transfusion. 1992;32:601-606.
13. Sazama K. Reports of 355 transfusion-associated deaths: 1976 through 1985. Transfusion.
1990;30:583-590.
14. Jensen N, Crosson J. An automated system for bedside verification of the match between
patient identification and blood unit identification. Transfusion. 1996;36:216-221.
15. Wenz B, Burns E. Improvement in transfusion safety using a new blood unit and patient
identification system as part of safe transfusion practice. Transfusion. 1991;31:401-403.
16. Cox C, Enno A, Deveridge S, Seldon M, Richards R, Martens V, et al. Remote electronic
blood release system. Transfusion. 1997;37:960-964.
17. Aller R, Freidman W. Rapid accuate entry of microbiology results. Arch Pathol Lab Med.
1996;120:57-61.
18. Willard K, Stanholtzer C. User interface reengineering: Innovative applications of bar
coding in a clinical microbiology laboratory. Arch Pathol Lab Med. 1995;119:706-712.
19. Bates D. Using information technology to reduce rates of medication errors in hospitals.
BMJ. 2000;320:788-791.
20. Meyer G, Brandell R, Smith J, Milewski F, Brucker J, P, Coniglio M. Use of bar codes in
inpatient drug distribution. Am J Hosp Pharm. 1991;48:953-966.
21. Gebhart F. VA slashed drug erros via bar-coding. Drug Topics. 1999; 143: 44.
22. Mudie D. Bar coding in a medical records department: a case study. Hosp Technol Ser.
1988;7:1-15.
493

23. Arenson RL, London JW. Comprehensive analysis of a Radiology Operations Management
computer system. Radiology. 1979;133:355-362.
24. Connelly A. Bar codes help control incomplete medical records. J Am Med Rec Assoc.
1987;58:48-49.
25. Burt TW, Bock HC. Computerized mega code recording. Ann Emerg Med. 1988;17:339-
342.
26. Bock HC. Field verification methodology using bar coding to record data. Ann Emerg Med.
1993;22:75-8; discussion 78-79.
27. Chua R, Cordell W, Ernsting K, Bock H, Nyhuis A. Accuracy of bar codes versus
handwriting for recording trauma recuscitation events. Ann Emerg Med. 1993;22:1545-
1550.
28. Kanmaz T, Haupt B, Peterson A. Comparison of manual and bar-code systems for
documenting pharmacists' interventions. Am J Health-Syst Pharm. 1997;54:1623-1626.
29. Zarowitz B, Petitta A, Mlynarek M, Touchette M, Peters M, Long P, et al. Bar-code
technology applied to drug use evaluation. Am J Hosp Pharm. 1993;50:953-959.
30. Scott M, McElnay J, Burnett K. Using bar-code technology to capture clinical intervention
data in a hospital with a stand-alone pharmacy computer system. Am J Health-Syst Pharm.
1996;53:651-654.
31. Renner S, Howanitz P, Bachner P. Wristband identification error reporting in 712
hospitals. Arch Pathol Lab Med. 1993;117:573-577.
32. Feldman S, Roblin D. Medical accidents in hospital care: applications of failure analysis to
hospital quality appraisal. Jt Comm J Qual Improv. 1997;23:567-580.
33. Schreiber G, Busch M, Kleinman S, Korelitz J. The risk of transfusion-transmitted viral
infections. The Retrovirus Epidemiology Donor Study. N Engl J Med. 1996;334:1685-
1690.
34. Lackritz J, Satten G, Alberle-Grasse J, al e. Estimated risk of transmission of the human
inmmunodeficiency virus by screened blood in the United States. N Engl J Med.
1995;333:1721-1725.
35. American Society of Hospital-System Pharmacists. National survey of pharmacy practice
in acute care settings: dispensing and administration - 1999. Am J Health-Syst Pharm.
2000;57:1759-1775.
36. Lau F, Wong R, Chui C, Ng E, Cheng G. Improvement in transfusion safety using a
specially designed transfusion wristband. Transfusion Medicine. 2000;10:121-124.
37. Amber R, Everett V. Emergency department patient tracking: A cost-effective system
using bar code technology. J Emerg Nurs. 1996;22:190-195.
38. Davis NM. Initiative for reducing medication errors: the time is now. Am J Hosp Pharm.
2000;57:1488.
39. American Society of Hospital-System Pharmacists. ASHP policy recommendation. Am J
Health Syst Pharm. 2001;58:603-606.
Subchapter 43.2. Strategies to Avoid Wrong-Site Surgery
Background
Operating on the wrong site or body part represents a potentially devastating event for all
parties involved. Cases of “wrong-site surgery” frequently attract considerable media attention 1-4
and foment malpractice lawsuits. Claims for wrong-site orthopedic surgeries result in indemnity
494

payments in 84% of cases, compared with only 30% of orthopedic claims overall. 5,6 Although
orthopedics represents the largest source of legal claims, the Physician’s Insurance Association
of America (PIAA) has handled wrong-site surgery litigation from the entire range of surgical
specialties and subspecialties. 6 Common factors identified in wrong-site surgery include the
involvement of multiple surgeons on a case, the performance of multiple procedures during a
single trip to the operating room, unusual time constraints, and unusual anatomy or patient
characteristics, such as physical deformity or morbid obesity. 7,8
Based upon careful review of 43 cases reported through its Sentinel Event Policy 9 over a
3-year period, the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO)
issued the following recommendations for avoiding wrong-site surgery: 7,8
• Mark the operative site and involve the patient in this process
• Require oral verification of the correct site in the operating room by each member of
the surgical team
• Follow a verification checklist that includes all documents and medical records
referencing the intended operative procedure and site
• Directly involve the operating surgeon in the informed consent process
• Engage in ongoing monitoring to ensure verification procedures are followed
Among these recommendations, marking the operative site has received the most
attention and is the focus of this chapter.
Practice Description
In 1998, the American Academy of Orthopaedic Surgeons endorsed a program of
preoperative surgical site identification called “Sign your Site,” modeled on the “Operate
through your initials” campaign instituted by the Canadian Orthopaedic Association from 1994-
96. 5,10 Both organizations recommend that the operating surgeon initial the intended operative
site, using a permanent marker, during a scheduled preoperative visit. For spinal surgery, the
recommendations additionally endorse the use of intra-operative x-rays for localization of the
pathologic spinal level before proceeding with the procedure. Many surgeons already employ
their own techniques for surgical site identification such as marking an “X” on the operative site
or marking “No” on the wrong limb. 5,11,12 While these practices are commendable, they have
theoretical drawbacks including lack of standardization across operating rooms and institutions.
These alternative strategies will not be evaluated further in this chapter.
Prevalence and Severity of the Target Safety Problem
From January 1995 to March 2001, JCAHO reviewed voluntary reports of 1152 “sentinel
events.” Wrong-site surgery accounted for 114 (9.9%) of these reports and included procedures
in neurosurgery, urology, orthopedics, and vascular surgery. 13 Despite the high profile of
JCAHO’s Sentinel Event Policy, 9 under-reporting by health care organizations almost certainly
affects these statistics. Only 66% of the 1152 total events were “self-reported” by the institutions
involved. The remainder came from patient complaints, media stories and other sources. 13 In
fact, using a mandatory reporting system, the New York State Department of Health received 46
reports of wrong-site surgery from April 1, 1998 through March 31, 2000 4 (F. Smith, personal
communication, May 2001), compared with the 114 cases JCAHO received nationally over a
495

period 3 times longer. 13 This suggests that voluntary incident reporting may underestimate the
true incidence by a factor of 20 or greater. 14 †
The PIAA reviewed claims data from 22 malpractice carriers representing 110,000
physicians from 1985 to 1995. 5 These claims included 331 cases of wrong-site surgery. The
complete PIAA database documents almost 1000 closed malpractice claims involving wrong-site
surgery. 6 However, this figure also underestimates the prevalence of wrong-site surgery, as every
case does not result in a claim. Most wrong-site surgeries involve relatively minor procedures
such as arthroscopy, 10,15 rather than limb amputations or major neurosurgical procedures.
Consequently sequelae are minimal. The State Volunteer Mutual Insurance Company
(Tennessee) released a series of 37 wrong-site surgery claims from 1977 to 1997. 15 Performing
the correct procedure on the wrong side constituted the most common error (eg, arthroscopic
knee surgery on the wrong knee in 15 of the 37 cases). Twenty-six of the patients experienced no
sequelae beyond a scar, and only three patients suffered permanent disability. Given the rarity of
significant harm, estimates of the incidence of wrong-site surgery derived from litigation data
likely underestimate the true prevalence of this problem, as do estimates based on incident
reports.
Opportunities for Impact
Some surgeons have developed their own methods for surgical site identification, 11,12 but
routine preoperative evaluation and marking of the intended surgical site by the attending
surgeon has yet to become standard practice in orthopedics or any surgical specialty. One year
after its "Sign Your Site" campaign, the American Academy of Orthopaedic Surgeons surveyed
its membership. Among the 2000 orthopedic surgeons surveyed, 77% responded that the idea
was a good one, but only 40% stated that they complied with the recommended practice. Only
one-third stated that their principal hospitals had a “Sign Your Site” or similar program in place,
although many anticipated initiation of one in response to the campaign. 16,17
Study Designs
The published literature includes no studies in which the adoption of a practice related to
surgical site identification is analyzed in the setting of a controlled observational design or
clinical trial. One report described the experience of 4 orthopedic surgeons in private practice, 18
but included no comparable observations from a control group. The experience of the Canadian
Orthopaedic Association remains unpublished, and the observed effect is based entirely on
litigation statistics 10 (B. Lewis, personal communication, March 2001).
Study Outcomes
Given the egregious and distressing nature of wrong-site surgery, the error itself
represents the outcome of interest, regardless of the clinical outcome. Unfortunately, in the
absence of observational studies that include controls, the number of malpractice claims for
wrong-site surgery represents the most widely cited outcome.
† Mandatory New York reporting system (NYPORTS) documented 46 events in 24 months.
New York State has a population roughly 1/15 th of the entire country. Thus, had JCAHO
captured wrong-site surgeries with the same sensitivity, one would expect 2484 wrong site
surgeries to have been reported during the 87 months covered by the JCAHO sentinel event
policy. This figure is 21 times the actual figure of 114 cases.
496

Evidence for Effectiveness of the Practice
The Canadian Medical Protective Association reported a baseline level of litigation for
wrong-site surgery at 7% of all orthopedic surgery settlements before 1994 when the Canadian
Orthopaedic Association instituted their “Operate Through Your Initials” policy. 10 Currently,
there are no known wrong-site surgery claims against orthopedic surgeons in Canada. (B. Lewis,
personal communication, March 2001) Interpreting the difference in the rates of litigation of a
rare occurrence is difficult, however, especially without an accurate estimate of the denominator
(ie, the total number of relevant procedures performed during the time periods involved).
Moreover, the degree to which Canadian surgeons complied with the policy is unknown. As
mentioned above, only 40% of responding American orthopedists reported adoption of
preoperative site identification in routine practice. 16
In a North Dakota private practice that used preoperative site identification with indelible
ink, there was one incidence of wrong-site surgery (pinning of the wrong phalanx in a hand) in
15,987 consecutive cases. 18 Even assuming that the sole detected case represents the only wrongsite
surgery in this sample, interpreting this low event rate is impossible without a control group.
Comparing this result with national data is also problematic. National data on the number of
orthopedics procedures performed each year might generate an estimate of an appropriate
“denominator,” but the nationwide “numerator” is unknown. Because we do not know the extent
to which incident reporting and malpractice litigation underestimate the incidence of wrong-site
surgery, we cannot accurately estimate the baseline rate of wrong-site surgeries for comparison
with results of a case series such as the North Dakota report cited above. 18
Potential for Harm
Some surgeons may worry that marking the surgical site increases the risk of
contamination, but this concern appears unwarranted. 15,18 More concerning is the potential harm
that may arise from confusion caused by practice variability in “signing the site.” Although the
original recommendations called for surgeons to initial the intended operative site, some
surgeons and hospitals mark the site with an “X.” 15,16 Still others use an “X” or “No” to mark the
limb or site that should not be operated upon. 11,12 In addition, there are reports of patients
crossing their legs before the ink is dry and producing an identical mark on the contralateral
knee, thus subverting the intended effect of the intervention. 10 Confusion may also ensue if
operating room personnel cover the mark with surgical drapes prior to the start of the surgery. 10
Costs and Implementation
The costs of marking a surgical site are negligible. Marking procedures that require the
presence of the surgeon in the preoperative area prior to the initiation of anesthesia may require a
culture shift among surgeons and involve the costs of surgeons’ time. Implementation strategies
designed to more efficiently utilize the operating surgeon’s time could be designed. For
hospitalized patients, implementation might involve the operating surgeon initialing the intended
operative site at the time consent is obtained, thus requiring that the physician be present at the
time of consent. The nurse/anesthetist, anesthesiologist or other responsible party in the preop
area would then be required to contact the operating surgeon only in cases where the operative
site has not already been initialed.
497

Comment
While “signing the site” represents a low-tech solution with high face validity, no
evidence supports a particular version of this practice. Additionally, the existence of different
versions of the “signing the site” practice may cause confusion to the point of increasing the
likelihood of error. Strategies that focus only on a single aspect of the identification problem,
without considering the preoperative and operative processes as a whole may fail to avert error.
For instance, protocols that rely on review of key preoperative x-rays in the operating room
creates a new mandate to ensure that the correct patient’s x-rays are brought to the operating
room. 6
Practices to successfully reduce (and eventually eliminate) wrong-site surgeries will
likely combine a standard method of marking the intended site with collaborative protocols for
verification of the intended procedure and operative site by all members of the operating room
staff. 7 Straightforward as each of these processes may appear, successful implementation may
require substantial investments of time and resources for protocol development and team
training. Whatever multifaceted practice is developed should be implemented within the setting
of a planned observational study.
References
1. Patient's son says family was seeking the best care (family of Rajeswari Ayyappan, who
had the wrong side of his brain operated on at Memorial Sloan-Kettering Cancer Center,
discusses his health). New York Times. Nov 16 1995: A16(N), B4(L).
2. Olson W. A story that doesn't have a leg to stand on (amputation of man's left leg rather
than right one may have been medically justified). Wall Street Journal. March 27 1995:
A18(W), A20(E).
3. Altman LK. State fines Sloan-Kettering for an error in brain surgery (New York State fines
Memorial Sloan-Kettering Cancer Center; surgeon operates on wrong side of patients
brain). New York Times. March 9 1996: 27(L).
4. Steinhauer J. So, the brain tumor's on the left, right? (Seeking ways to reduce mix-ups in
the operating room; better communication is one remedy, medical experts say). New York
Times. April 1 2001: 23(N), 27(L).
5. American Academy of Orthopaedic Surgeons. Advisory statement on wrong-site surgery.
Available at: http://www.aaos.org/wordhtml/papers/advistmt/wrong.htm. Accessed April
16, 2001.
6. Prager LO. Sign here. Surgeons put their signatures on patients' operative sites in an effort
to eliminate the change of wrong-site surgeries. Am Med News. October 12 1998: 13-14.
7. The Joint Commission on Accreditation of Healthcare Organizations. Sentinel Event Alert.
Lessons Learned: Wrong Site Surgery. Available at: http://www.jcaho.org/edu_pub/
sealert/sea6.html. Accessed March 30, 2001.
8. Lessons learned: sentinel event trends in wrong-site surgery. Jt Comm Perspect.
2000;20:14.
9. Sentinel events: approaches to error reduction and prevention. Jt Comm J Qual Improv.
1998;24:175-86.
10. Position paper on wrong-sided surgery in orthopaedics. Winnepeg, Manitoba: Canadian
Orthopaedic Association Committee on Practice and Economics; 1994.
11. Lubicky JP. Wrong-site surgery. J Bone Joint Surg Am. 1998;80:1398.
498

12. Cowell HR. Wrong-site surgery. J Bone Joint Surg Am. 1998;80:463.
13. Joint Commission on Accreditation of Healthcare Organizations. Sentinel Event Statistics.
Available at: http://www.jcaho.org. Accessed April 16, 2001.
14. U.S. Census Bureau. United States Census 2000: Ranking Tables for States: Population in
2000 and Population Change from 1990 to 2000. Available at:
http://www.census.gov/population/www/cen2000/phc-t2.html. Accessed May 31, 2001.
15. American Academy of Orthopaedic Surgeons. Report of the task force on wrong-site
surgery. Available at: http://www.aaos.org/wordhtml/meded/tasksite/htm. Accessed March
15, 2001.
16. Risk Management Foundation of the Harvard Medical Institutions. Did Wrong-Site
Surgery Remedy Work? Available at: http://www.rmf.harvard.edu/publications/resource/
feb1999news/article2/. Accessed April 16, 2001.
17. The American Academy of Orthopedic Surgeons. Academy News. "Sign your site" gets
strong member support. Available at: http://www.aaos.org/wordhtml/99news/os3-sign.htm.
Accessed May 5, 2001.
18. Johnson PQ. Wrong site surgery in orthopaedics: analysis of 15,987 cases at The
Orthopaedic Instititute in Fargo. Paper presented at: American Academy of Orthopaedic
Surgeons 67th Annual Meeting; March 15, 2000, 2000; Los Angeles, CA.
499

500

Chapter 44. Crew Resource Management and its Applications in Medicine
Laura Pizzi, PharmD
Neil I. Goldfarb
David B. Nash, MD, MBA
Thomas Jefferson University School of Medicine
and Office of Health Policy & Clinical Outcomes
Background
Patient care, like other technically complex and high risk fields, is an interdependent
process carried out by teams of individuals with advanced technical training who have varying
roles and decision-making responsibilities. While technical training assures proficiency at
specific tasks, it does not address the potential for errors created by communicating and decision
making in dynamic environments. Experts in aviation have developed safety training focused on
effective team management, known as Crew Resource Management (CRM). Improvements in
the safety record of commercial aviation may be due, in part, to this training. 1 Over the past 10
years, lessons from aviation’s approach to team training have been applied to patient safety,
notably in intensive care unit (ICU) and anesthesia training. 2,3 This chapter reviews the literature
on Crew Resource Management, also known as Cockpit Resource Management, and describes
adaptations of this training framework to medicine.
Practice Description
Crew Resource Management in Aviation
Crew Resource Management has been widely used to improve the operation of flight
crews. The concept originated in 1979, in response to a NASA workshop that examined the role
that human error plays in air crashes. 4 CRM emphasizes the role of human factors in high-stress,
high-risk environments. John K. Lauber, a psychologist member of the National Transportation
Safety Board, defined CRM as “using all available sources—information, equipment, and
people—to achieve safe and efficient flight operations.” 5,6 CRM encompasses team training, as
well as simulation (also referred to as Line-Oriented Flight Training, or LOFT), interactive
group debriefings, and measurement and improvement of aircrew performance.
There is no universal CRM training program. The Federal Aviation Administration
(FAA) allows air carriers to customize their CRM programs to best suit the needs of individual
organizations. Therefore, training programs vary somewhat from carrier to carrier, making it
difficult to describe operational components. Furthermore, these programs continue to evolve as
aviation technology changes and more is learned about group dynamics.
One CRM model focuses on the elements of human effectiveness. 7 The 3 primary
components of effective crew management are safety, efficiency, and morale. Specific factors
related to aircrew performance are categorized, and serve as the basis for training and research.
These factors include materials, organization, individual, and group process variables associated
with performance. 8 Examples of outcomes that result from these input variables are safety,
efficiency, and customer satisfaction.
Subsequently, Helmreich and colleagues proposed a modified conceptual framework for
CRM, termed the “Error Troika,” 9 to display a hierarchy of 3 error countermeasures. At the first
501

level, CRM includes training on how to avoid errors. At the second level, potential errors are
“trapped” before they are committed. At the third level, mitigation of error consequences occurs.
From a practical standpoint, CRM programs typically include educating crews about the
limitations of human performance. 10 Trainees develop an understanding of cognitive errors, and
how stressors (such as fatigue, emergencies, and work overload) contribute to the occurrence of
errors. Multi-day CRM training programs typically require participants to assess personal and
peer behavior. Operational concepts stressed include inquiry, seeking relevant operational
information, advocacy, communicating proposed actions, conflict resolution and decision
making.
Prevalence and Severity of the Target Safety Problem
The field of aviation has a substantial history of collecting and analyzing safety-related
data. Historically, human error has caused or contributed to over 50% of aviation accidents. In an
analysis of 35,000 reports of incidents over 7.5 years, almost 50% resulted from a flight crew
error, and an additional 35% were attributed to air traffic controller error. 11 Root cause analyses
(Chapter 5) by safety experts have found that errors frequently occur because flight crews fail to
effectively manage the resources available to them (eg, fail to verify information when uncertain
about it, fail to plan for contingencies). 11 Naval aviation reports provide similar results, with one
study reporting 59% of “Class A mishaps” (serious consequences including fatality, destroyed
aircraft, and major injury) attributed to some degree to aircrew factors. 12 These and similar
analyses have catalyzed tailored prevention strategies including CRM for commercial aviation, 1
and “aircrew coordination training” (ACT) for Naval aviators. 12
Study Design and Outcomes
Measures
Although the most obvious and meaningful measure of CRM effectiveness would appear
to be airline accident or “near miss” rates, these objective measures have not been used in
commercial aviation studies. Helmreich suggests that it is not possible to use these measures,
because accident rates are very low and “near misses” are voluntarily reported. 10 Furthermore,
the content and structure of CRM training programs are variable. 10
In response, researchers have developed tools that assess the effectiveness of CRM in
other ways. These tools include attitudinal surveys and peer performance rating questionnaires,
including the NASA/University of Texas Line/LOS Checklist (LINE/LOS Checklist), 13 the
Cockpit Management Attitudes Questionnaire (CMAQ), and the Flight Management Attitudes
Questionnaire (FMAQ). The LINE/LOS Checklist is used to rate crew performance on critical
behaviors during specific segments of flight (eg, not rushing through briefing period, exhibiting
high levels of vigilance in both high and low workload conditions, etc). Ratings on each
behavioral element (ie, model for teamwork) range across 4 levels from poor to outstanding.
In contrast, CMAQ is used to evaluate the attitudes of crewmembers within and between
organizations, pre- and/or post-CRM training. Results are intended to serve as a proxy for
measuring crew process and performance. 14 The instrument has been validated by comparing
self-reported attitudes with performance ratings made by experienced Check Airmen, experts
trained in peer evaluation. 15 The FMAQ is a revised version of the CMAQ that was developed by
Helmreich and colleagues in response to attitudinal differences observed in flight crews from
different countries. 16
Observation of crew performance in simulated flights has also been used.
502

Representative Studies
Several studies have utilized proxy tools to test the effectiveness of CRM. 8,12,17 One study
by Helmreich and colleagues consisted of an assessment of the attitudes before versus after CRM
training (pre-test versus post-test). 17 Crew behaviors were noted by a trained observer using the
NASA/University of Texas Line/LOS Checklist. 18 More than 2000 line flights and LOFT
sessions were included in the analysis. Overall performance of crews was classified as “below
average,” “average,” or “above average” by Check Airmen and LOFT instructors.
As a result of the CRM training, the percentage of crews rated as “above average”
increased while the percent rated “below average” decreased. Performance ratings differed
between fleets and airlines. Superior pilots also shared many common attitudes, (for example,
they were aware of their personal limitations and diminished decision-making capacity during
emergencies). In addition, they encouraged crewmembers to question their decisions and actions,
were sensitive to the personal problems of other crewmembers, and recognized the need to
verbalize plans and to train other crewmembers.
A second study by Barker and colleagues compared the effectiveness of 17 CRM-trained
flight crews on military mission simulators, divided according to whether they were “fixed” or
“formed” crews. 8 Nine of the crews were defined as "fixed" since they had flown together for six
months or longer; the remaining 8 were “formed,” as they had flown together for less than 6
months. Each crew was asked to participate in a simulated mission. The first leg of the mission
was programmed to be problem-free, but the second leg required crews to address a safety issue.
As with the earlier study, crew behaviors were observed using the NASA/University of Texas
Line/LOS Checklist. Surprisingly, the formed crews committed fewer minor errors, but the
number of major errors did not differ significantly between the groups. 8 The authors concluded
that formed crews may experience less ineffective coordination than fixed crews, as the latter
might be more complacent from the routine of working together, but that further research was
needed.
A third study evaluated rates of naval aviation mishaps, and the role of CRM failures. 12
While the study primarily compared crew performance in 2 types of equipment, the aircrew
performance deficits were also compared with those seen during an earlier time period, prior to
implementation of aircrew coordinating training (ACT) programs (the specific form of CRM
used). Analysis of data from the post-ACT implementation period across the fleet revealed that
70% of human error mishaps were connected with aircrew factors, and that 56% of these resulted
from at least one CRM failure. The authors noted that these percentages were similar to those
reported in a separate study prior to ACT implementation. Because of a lack of controls or
adjustments for potential confounders, no conclusions about effectiveness of the ACT program
can be drawn. 12
503

Comparison to Medicine
Sexton and colleagues compared flight crews with operating room personnel on several
measures, including attitudes toward teamwork. 19 The study included more than 30,000 cockpit
crew members (captains, first officers, and second officers) and 1033 operating room personnel
(attending surgeons, attending anesthesiologists, surgical residents, anesthesia residents, surgical
nurses, anesthesia nurses). Data from the crew members were obtained from previous
administrations of the CMAQ and FMAQ to major airlines around the world (over a 15-year
period). The operating room participants were mailed an analogous questionnaire (CMAQ
modified 20 ), administered over a period of 3 years at 12 teaching and non-teaching hospitals in
the United States and abroad (Italy, Germany, Switzerland, and Israel).
The level of teamwork perceived by attending surgeons compared with other operating
room staff differed markedly. A majority of surgical residents (73%) and attending surgeons
(64%) reported high levels of teamwork, but only 39% of attending anesthesiologists, 28% of
surgical nurses, 25% of anesthesia nurses, and 10% of anesthesia residents reported high levels
of teamwork. A bare majority (55%) of attending surgeons rejected steep hierarchies (determined
by whether they thought junior team members should question the decisions of senior team
members). In contrast, 94% of airline crew members preferred flat hierarchies.
It was also noted that medical participants were far more likely to agree with the
statement “Even when fatigued, I perform effectively during critical times.” Seventy percent of
attending surgeons agreed with this statement, as well as 56% of surgical residents, 60% of
surgical nurses, 57% of anesthesia residents, 55% of anesthesia nurses, and 47% of attending
anesthesiologists. In contrast, 26% of pilots agreed with this statement.
Applications of CRM Principles in Medicine
The Sexton study and other analyses suggest that safety-related behaviors that have been
applied and studied extensively in the aviation industry may also be relevant in health care. We
identified CRM applications in several dynamic decision-making health care environments: the
operating room, labor and delivery, and the emergency room. 3,21,22 In addition, Gaba has noted
that some other domains (eg, cardiac arrest response teams) that have active simulation training
are currently incorporating a broader range of CRM-like training methods. 23 (Simulators are
covered in more detail in Chapter 45).
Practice Description
Crew Resource Management in Health Care Settings
As with aviation, the medical application of CRM has required tailoring of training
approaches to mirror the areas in which human factors contribute to mishaps. In anesthesiology
65-70% of safety problems (accidents or incidents) have been attributed at least in part to human
error. In response, several anesthesiologists from the VA Palo Alto Health Care System and
Stanford University, with funding from the Anesthesia Patient Safety Foundation, developed
Anesthesia Crisis Resource Management (ACRM), modeled on CRM. 3,23 The original
demonstration courses consisted of didactic instruction, videotape of a reenactment of an
aviation disaster, videotape of an actual anesthetic mishap, simulation training and a debriefing
session. Ongoing courses include the use of a textbook that catalogues 83 critical events (eg,
acute hemorrhage, bronchospasm, seizures), and approaches to managing them. 24 Currently,
there are 3 ACRM courses offering progressively more challenging material, a Working Group
504

on Crisis Management Training in Health Care formed by the developers of ACRM that has
initiated formal ACRM instructor training, and thoughtful consideration of pragmatic approaches
to evaluating ACRM. 23
Helmreich and Schaefer have also advanced CRM theory in the operating room
environment by adapting their model of team performance. This framework describes the team
performance inputs that are critical to essential team functions that in turn lead to desired
outcomes (defined as patient well-being). Examples of inputs include individual aptitudes,
physical environment, and culture (professional, organizational, and national). Performance
functions consist of team formation and management, surgical procedures, communications,
decision processes, and situational awareness. 18
Another application of CRM to the health environment is the MedTeams behavior-based
teamwork system, developed by Dynamics Research Corporation, and sponsored by the Army
Research Laboratory. It aims to adapt research in team performance and training from military
helicopter aviation to emergency medicine. 22,25,26 Thus far, specific applications have been
developed for Emergency Department care and labor and delivery units. 27 The system is
implemented through courses and assessment tools. The Emergency Team Coordination Course
(ETCC) includes 5 team dimensions or goals (ie, maintain team structure and climate, facilitate
planning and problem-solving, enhance communication among team members, facilitate
workload management, improve team-building skills). Each goal is tied to specific teamwork
tasks. For example, tasks for the first goal (maintain team structure and climate) include
“establish team leader,” “form the team,” “set team goals,” and “assign roles and
responsibilities.” 22 Like the CRM approach to the “Error Troika,” the MedTeams approach is
based on avoiding errors, trapping them as they occur, and mitigating the consequences of actual
errors. Principles underlying the MedTeams approach include:
• Team responsibility for patients
• A belief in clinician fallibility
• Peer monitoring
• Team member awareness of patient status, team member status and
institutional resources
Peer monitoring is a fundamental component of the MedTeams system, as well as a
feature of the ACRM and aviation field’s CRM approaches. Along with his or her clinical
responsibilities, each team member undertakes the intermittent process of peer monitoring or
"check" actions, engaging in this check cycle as frequently as possible. The teamwork check
cycle begins with each team member monitoring his or her own situation awareness and crossmonitoring
the actions of other teammates. If during the monitoring mode the monitoring
teammate observes a suspected error in progress, that individual intervenes with a direct question
or offer of information. The erring teammate may then acknowledge the lapse, correct it and
continue working. Alternatively, the monitoring teammate may have lost situation awareness.
The non-erring monitored colleague can then provide feedback to correct the peer's situation
awareness. If team members are in strong disagreement about how patient care should proceed,
advocacy, assertion and perhaps third-party involvement may be used to resolve the situation.
Over time, the check cycle becomes habitual, resulting in hundreds of team checks daily, all with
the potential to break the error chain.
Recently, ACRM has been extended to a full-day course on neonatal resuscitation
training for neonatologists and pediatricians. 21 Called “NeoSim”, the course combines traditional
505

training methods with reviews of literature and didactic instruction with simulation. Debriefing
follows, using videotape of the simulation. As with the other examples, the emphasis of teaching
behavioral teamwork skills along with technical content is the hallmark of CRM interventions in
health care.
Evidence for Effectiveness of the Practice
The most thoroughly studied of the medical CRM applications is ACRM, although as
with the aviation examples, rigorous evaluations are challenging to design. Few studies utilize a
control group, although researchers have reported assessment methods for determining both
technical and behavioral performance from simulator videotapes 28 (see also Chapter 45). A
before-after ACRM training analysis (Level 3 Study Design) of trainees’ knowledge-base (Level
3 Outcome) for crisis yielded mixed results. 3 The average score on the post-test was significantly
greater than the pre-test for one trainee class, composed mostly of residents. For the other class
of experienced anesthesiologists, the test scores did not change and were at the same level as the
post-test scores of class of residents. Subjective data evaluating the course indicated that trainees
“uniformly felt that the ACRM course was an intense, superior form of training related to an
important, but inadequately taught, component of anesthesia practice.” 3 Another study of ACRM
at Harvard also found that participants rated the course favorably, with over 80% responding that
they felt the course should be taken every 24 months or less. 29
As with aviation, the incremental value of this form of training is difficult to link to
improvements in teamwork performance and better safety records. At the time of this literature
assessment, there were no published data to describe the effects on medical error rates of the
MedTeams approach. The NeoSim course participants provided positive responses to openended
questions about their satisfaction with the course. 21
Costs and Implementation
Helmreich has noted some of the limitations associated with CRM. 10 At this time, the
evidence connecting CRM approaches to improving patient safety does not exist,
notwithstanding the face validity of the approach. Nevertheless, a long history of variants of the
approach offers health care a reasonable foundation from which to draw practical and evidencedbased
resources 30 for further development and adaptation of CRM, as well as measurement
methods to ascertain its effectiveness.
At a minimum, implementation of the CRM approach in health care settings requires
customization of tools and techniques for each specific care venue, as is illustrated by
adaptations implemented thus far. This customization comes at considerable cost and cannot be
expected to immediately reap safety benefits. At the time of this review, approximate costs for
implementing the MedTeams system ranged from $15,000-$35,000, with additional costs for
ongoing activities (such as continuing education) that ranged from $8,000-$20,000 (R. Simon,
personal communication, April 2001). Similarly, marginal costs for CRM-like training based on
the ACRM experience are estimated at $800 to $2,000 per participant per day (D. M. Gaba,
personal communication, June 2001). These costs do not include the overhead of starting a
program (eg, simulator investment, training instructors), nor do they factor in the cost of reduced
practice work hours, if these are above those devoted to current training time.
Cultural shifts in medicine are also necessary if the CRM approach is truly to take root.
CRM applications are relatively novel in health care, a field in which professional training and
education have traditionally focused on developing technical proficiency rather than facilitating
human interaction. Although communication and decision making are central to medical
506

practice, relatively little about this topic has appeared in medical literature, despite the fact that
information flow is critical, particularly in high acuity venues such as the operating room and
emergency departments. Ideas must be elicited, debated and evaluated without discrimination
based on the status of the staff person offering the information. 31
Paradoxically, the attachment to hierarchy may be the reason that a small percentage of
participants can be expected to reject CRM training. Research has shown that some resistance is
rooted in personality characteristics. Crew members lacking in achievement motivation and
interpersonal skills are more likely to reject the training. Additionally, CRM practices decay over
time, even with repeated training, requiring continuing expenditures to retain the hoped-for
gains. 10
Comment
CRM has evolved in the airline industry for more than 20 years, and has been extensively
applied during the past decade. Although no definitive data link CRM to decreased aviation error
per se, the industry has accepted the face validity of the practice, and it is now an integral part of
training. Over time, survey data from thousands of civilian and military participants in the
United States and abroad has been accrued. These data indicate that most flight crew members
accept CRM training, and find it both relevant and useful. 7
The studies reviewed provide some support for the notion that CRM is worth further
investigation in health care. However, it cannot yet be concluded that CRM is a practice that can
reduce medical errors. Additional research in this area is warranted, although measurement and
study design are particularly challenging. As noted, although the evidence from aviation, where
CRM is well-established, has shortcomings, data are easier to capture since CRM assessments
can be grafted onto mandatory, ongoing, yearly pilot assessments conducted both in simulators
and in real flights (D. M. Gaba, personal communication, June 2001). In medicine, where
ongoing assessments are not the norm, capturing relevant data will be more difficult logistically
and much more expensive than in aviation. Consequently, and for the analogous reasons that
aviation has adopted CRM based on face validity, health care decision makers may wish to
consider face validity in lieu of massive research investments.
Nonetheless, evaluations of CRM that focus on intermediate outcomes (eg, trainee
performance) are feasible, and instructive for optimizing components of CRM programs. CRM
design and evaluation resources are becoming more widely available, 30 and health care should
continue to consult these sources, building on the advances made in other fields.
Acknowledgements
The authors are grateful to Robert Simon, Chief Scientist at Dynamics Research
Corporation, for providing details on the MedTeams approach.
References
1. Helmreich RL. On error management: lessons from aviation. BMJ. 2000; 320: 781-5.
2. Shortell SM, Zimmerman JE, Rousseau DM, Gillies RR, Wagner DP, Draper EA, et al.
The performance of intensive care units: does good management make a difference? Med
Care. 1994; 32: 508-25.
3. Howard SK, Gaba DM, Fish KJ, Yang G, Sarnquist FH. Anesthesia crisis resource
management training: teaching anesthesiologists to handle critical incidents. Aviat Space
Environ Med. 1992; 63: 763-70.
507

4. Cooper GE, White MD, Lauber JK. Resource management on the flightdeck: proceedings
of a NASA/ Industry Workshop. Moffett Field, Calif: NASA - Ames Research Center; 1980.
NASA Conference Publication No. CP-2120.
5. Lauber JK. Cockpit resource management: background and overview. In: Orlady HW,
Foushee HC, eds. Cockpit resource management training: proceedings of the NASA/MAC
workshop. Moffett Field, Calif: NASA - Ames Research Center. NASA Conference
Publication No. 2455.
6. Wiener EL, Kanki BG, Helmreich RL. Cockpit resource management. San Diego, Calif:
Academic Press, Inc.; 1993.
7. Helmreich RL, Foushee HC. Why crew resource management? Empirical and theoretical
bases of human factors in training and aviation. In: Wiener E, Kanki BG, Helmreich RL,
eds. Cockpit resource management. San Diego, Calif: Academic Press; 1993: 3-45.
8. Barker JM, Clothier CC, Woody JR, McKinney EH, Jr., Brown JL. Crew resource
management: a simulator study comparing fixed versus formed aircrews. Aviat Space
Environ Med. 1996; 67: 3-7.
9. Helmreich RL, Merritt AC. Cultural issues in crew resource management. Conference
presentation at the ICAO Global Human Factors Seminar; April, 1996; Auckland, New
Zealand.
10. The evolution of crew resource management training in commercial aviation. Available at:
http://www.psy.utexas.edu/psy/helmreich/Evolution_IJAP_for_Dist.htm. Accessed June
18, 2001.
11. Billings CE, Reynard WD. Human factors in aircraft incidents: results of a 7-year study.
Aviat Space Environ Med. 1984; 55: 960-5.
12. Wiegmann DA, Shappell SA. Human error and crew resource management failures in
Naval aviation mishaps: a review of U.S. Naval Safety Center data, 1990-96. Aviat Space
Environ Med. 1999; 70: 1147-51.
13. Helmreich RL, Wilhelm JA, Kello JE, Taggart WR, Butler RE. Reinforcing and evaluating
crew resource management: Evaluator/LOS instructor reference manual. Austin, Tex:
NASA - University of Texas at Austin; 1990. Technical Manual 90-2.
14. Gregorich SE, Helmrich RL, Wilhelm JA. The structure of cockpit-management attitudes.
J Appl Psychol. 1990; 75: 682-90.
15. Helmreich RL, Foushee BR, Benson R, Russini W. Cockpit Resource Management:
Exploring the attitude-performance linkage. Aviat Space Environ Med. 1986; 57: 1198-200.
16. Helmreich RL, Merritt AC, Sherman PJ, Gregorich SE, Wiener EL. The flight management
attitudes questionnaire. Austin, Tex: NASA/University of Texas/FAA; 1993. Technical
Report 93-5.
17. Helmreich RL, Wilhelm JA, Gregorich SE, Chidester TR. Preliminary results from the
evaluation of cockpit resource management training: performance ratings of flight crews.
Aviat Space Environ Med. 1990; 61: 576-9.
18. Helmreich RL, Schaefer HG. Team performance in the operating room. In: Bogner MS, ed.
Human error in medicine. Hillside, NJ: Lawrence Erlbaum; 1998.
19. Sexton JB, Thomas EJ, Helmreich RL. Error, stress, and teamwork in medicine and
aviation: cross sectional surveys. BMJ. 2000; 320: 745-9.
20. Helmreich RL, Merritt AC. Culture at work in aviation and medicine: national,
organizational, and professional influences. Hants, England: Ashgate Publishing Limited;
1998.
508

21. Halamek LP, Kaegi DM, Gaba DM, Sowb YA, Smith BC, Smith BE, et al. Time for a new
paradigm in pediatric medical education: teaching neonatal resuscitation in a simulated
delivery room environment. Pediatrics. 2000; 106: E45.
22. Risser DT, Rice MM, Salisbury ML, Simon R, Jay GD, Berns SD. The potential for
improved teamwork to reduce medical errors in the emergency department. The MedTeams
Research Consortium. Ann Emerg Med. 1999; 34: 373-83.
23. Gaba DM, Howard SK, Fish KJ, Yasser SA. Simulation-based training in Anesthesia Crisis
Resource Management (ACRM): a decade of experience. In: Simulation & Gaming. Vol
32: Sage Publications, Inc. In Press.
24. Gaba DM, Fish KJ, Howard SK. Crisis Management in Anesthesiology.: Churchill
Livingstone Inc.; 1994.
25. Small SD, Wuerz RC, Simon R, Shapiro N, Conn A, Setnik G. Demonstration of highfidelity
simulation team training for emergency medicine. Acad Emerg Med. 1999; 6: 312-
23.
26. Dynamics Research Corporation. Training analysis and delivery. Available at:
http://www.drc.com/TrainingAnalysis/medteams.htm. Accessed June, 2001.
27. Spath PL, ed. Error reduction in health care: a systems approach to improving patient
safety. San Francisco, Calif: Jossey-Bass Publishers; 1999.
28. Gaba DM, Howard SK, Flanagan B, Smith BE, Fish KJ, Botney R. Assessment of clinical
performance during simulated crises using both technical and behavioral ratings.
Anesthesiology. 1998; 89: 8-18.
29. Holzman RS, Cooper JB, Gaba DM, Philip JH, Small SD, Feinstein D. Anesthesia crisis
resource management: real-life simulation training in operating room crises. J Clin Anesth.
1995; 7: 675-87.
30. Salas E, Rhodenizer L, Bowers CA. The design and delivery of crew resource management
training: exploiting available resources. Hum Factors. 2000; 42: 490-511.
31. Schenkel S. Promoting patient safety and preventing medical error in emergency
departments. Acad Emerg Med. 2000; 7: 1204-22.
509

510

Chapter 45. Simulator-Based Training and Patient Safety
Ashish K. Jha, MD
University of California, San Francisco School of Medicine
Bradford W. Duncan, MD
Stanford University School of Medicine
David W. Bates, MD, MSc
Harvard Medical School
Background
For a number of years, simulators have been used in aviation, nuclear power, military
flight operations and other industries as a training tool and method to assess performance. Their
use is nearly universal in high reliability organizations. 1 Recently the use of simulation in
medicine has increased markedly, in part due to greater awareness of the importance of patient
safety.
Defined broadly, a simulator replicates a task environment with sufficient realism to
serve a desired purpose. 1 In medical training, simulators can substitute for actual patients and can
be as simple as utilizing pigs’ feet to practice suturing, or as complex as virtual reality machines
and re-creations of actual clinical environments for surgeons, radiologists and anesthesiologists.
In a general sense, they improve patient safety by allowing physicians to become better trained
without putting patients at risk. For example, in a randomized controlled trial, Peugnet and
colleagues used a virtual reality simulator to train physicians to perform retinal photocoagulation
and found that surgeons who trained with the simulator performed the procedure as well as those
who trained with patients. 2 Gaba lists several other advantages of simulation, among them 3 :
• Presentation of uncommon but critical scenarios in which a rapid response is
needed (eg, malignant hyperthermia, which occurs once in every 40,000
anesthesia cases). 4 To conduct systematic training about managing such
critical events there is little alternative but to use simulation.
• Errors can be allowed to occur and reach their conclusion—in real life a more
capable clinician would have to intervene—so participants can see the results
of their decisions and actions.
• With mannequin-based simulators clinicians can use actual medical
equipment, exposing limitations in the human-machine interface.
• Complete interpersonal interactions with other clinical staff can be explored
and training in teamwork, leadership, and communication provided. In a
number of medical fields, simulation has been used in crew resource
management (CRM) training 1 (see Chapter 44), where the focus is on
behavioral skills such as inter-team communication during critical incidents.
Human error is a leading cause of adverse anesthesia events that lead to poor patient
outcomes 5 and may play an important role in surgical errors as well. 6 However, it may be
difficult to demonstrate improved patient outcomes from simulation because adverse events are
unusual and there are an extreme number of potential confounders. Given these difficulties,
performance has been used as a surrogate outcome, despite concerns that it may be a less than
perfect measure. It is yet unclear which attributes of performance matter most to patient
outcomes. Additionally, the methods (eg, reliability and consistency) and timing (eg, duration of
511

follow-up) by which performance is measured also remain ill-defined. 1 Studies of the
effectiveness of simulators often are limited in that they measure performance using the same
training simulator, which may favor those who have trained on the simulator itself. In other
words, seemingly improved performance may not translate to actual patient care. Of the studies
that have extended laboratory simulation to patient care, few have evaluated the impact on
medical error or established a clear link between simulator training and patient outcomes.
This chapter reviews the evidence regarding the use of simulators in the training and
ongoing education of health care providers as a patient safety measure. Not included in this
review is the use of simulators in the planning or preparation of diagnostic or therapeutic
interventions in specific patients (eg, calculation of radiotherapy doses using computer
simulations, planning a surgical procedure for a specific patient using 3-D simulations based on
the anatomical data from the actual patient). While these kinds of simulation clearly decrease the
morbidity and mortality associated with their particular procedures, they are omitted here
because they focus on unique characteristics of individual patients that may not be generalizable
to other patients. Other uses of simulators beyond the scope of this discussion include simulators
to measure worker proficiency, 7-10 to identify areas for educational intervention, and to target
improvements in training. 11,12
Simulators in Anesthesia
Patient simulators have been most widely studied in anesthesia where human error may
account for over 80% of critical incidents. 5 Simulators range from simple mannequins to highfidelity
simulators that recreate the operating room experience. According to one study, 71% of
medical schools in Canada, the United Kingdom and other western nations used mannequins or
some other form of simulator to teach anesthesia to medical students. 13 There is a growing body
of literature on the different roles that simulators can play in anesthesia training. 14
Studies have found that simulators can effectively identify errors and appropriateness of
decision making in anesthesia. For example, during 19 comprehensive anesthesia simulations,
DeAnda et al documented 132 unplanned incidents, of which 87 (66%) were due to human error
and 32 (27%) were considered critical incidents. 15 Schwid and O’Donnell have also used
simulators to document the type of errors that anesthesiologists make in critical incidents,
finding errors in monitor usage (37%), airway management (17%), ventilator management
(13%), and drug administration (10%). 16 Gaba and colleagues studied both the appropriateness
of decisions and response time of anesthesia trainees to simulated critical incidents. 8 They found
great individual variability as well as variability by incident in the accuracy and timeliness of
response. Some simulated incidents, such as cardiac arrest, had major errors in management a
majority of the time. Based on these studies, patient simulators can be used to identify areas for
further education or training of anesthesia providers.
We identified 2 studies of the effect of simulators on anesthetists’ performance. Schwid
and colleagues studied the impact of a computer screen-based anesthesia simulator in a
randomized, controlled trial of 31 first-year anesthesia residents. 17 Residents that had trained on
the simulator with individualized debriefing responded better to critical events on a mannequinbased
simulator than those who received standard training without the simulator. Using a
randomized, controlled design, Chopra and colleagues studied management of simulated critical
situations by 28 resident or staff anesthesiologists. 18 The performance of subjects who trained on
the simulator was superior to that of subjects who did not receive that training.
Another setting where simulators may play an important role is in crew resource
management (CRM). Though establishing the effectiveness of simulation in CRM training may
512

e difficult, 19 initial work has been done on reliably and consistently rating performance. 10 CRM
is discussed further in Chapter 44.
Proficiency on a simulator does not ensure proficiency in clinical settings. Simulator
fidelity (ie, how accurately the simulator replicates reality) is imperfect. It is much more difficult
to “re-create” a human being than to do so for, say, an airplane. This limitation is illustrated by a
study conducted by Sayre and colleagues. They studied emergency medical technicians (EMT)
who learned intubation techniques on anesthesia mannequins. 20 After successfully intubating the
mannequins 10 times, they were permitted to intubate patients in the field, where their
proficiency was only 53%. Other factors can inhibit optimum learning using simulation or the
applicability of learning to real practice. Some participants may be more vigilant than usual
during simulator sessions. Others may be unable to “suspend disbelief,” may treat the simulation
only as a game, or act in a cavalier fashion, knowing that the simulator is not a real patient. 21
Refinement of simulators to make them more sophisticated and life-like may help to improve the
quality of the training that simulators can provide. Appropriate construction of curricula and
debriefings can also minimize the potential problems of simulation training.
Simulators in Radiology
The number of radiologic examinations that require sedation, analgesia, or contrast media
has increased rapidly in recent years. 22 Despite their rarity, serious medication reactions do occur
and require prompt, appropriate management. Some evidence of suboptimal management 23 has
prompted the creation of computer-based simulators to improve training in these areas. 24
Simulators have also been used to measure the effectiveness of strategies to teach trainees about
critical incidents, but studies have not reported the effectiveness of simulators as adjuncts to
training. Sica and colleagues developed a computer-based simulator that was used to study the
effectiveness of a lecture and videotape-based intervention on critical incidents for radiology
housestaff. 25 Those residents who underwent the intervention scored better on the simulator than
those that had only received basic, standard training. The authors concluded that the simulator
was an effective way of assessing the utility of the educational course.
Simulators in Surgery
As surgical technique and expertise has changed drastically over recent decades, some
methods used to train surgeons have evolved as well. Simulators in the surgical setting are aimed
at improving surgeons’ technical skills and dexterity. Training on simulators and virtual reality
machines, though still in its nascent stages, is becoming increasingly accepted. 26,27 Surgical
simulators have been developed for a variety of procedures: endovascular repair of abdominal
aortic aneurysms, 28 sinus surgery 29,30 gynecologic surgery, 31 orthopedic surgery, 32 prostatic
surgery, 33 amniocentesis procedures, 34 and oral surgery. 35 Nonetheless, many of these have yet to
be formally evaluated in terms of efficacy in improving physician performance in patient care.
We identified several studies that evaluated physician performance after training on a
surgical simulator. Derossis evaluated surgical residents and attendings in a randomized study
and found that those trained with a simulator had greater proficiency in suturing, transferring,
and mesh placement, when tested on the simulator, than did the control group. 36 They
subsequently found that when tested in vivo in pigs, surgeons (both attendings and residents)
who had been randomized to the simulator arm were more proficient at the same skills. 37 Scott
and colleagues studied the impact of a video-trainer on laparoscopic cholecystectomy skills. 38
They randomized surgical residents to training on a video-trainer versus no formal training over
513

a 30-day period. At the end of the period, when tested on pre-specified tasks on the video trainer,
those who trained on the video trainer did uniformly better.
Intuitively, improved technical skills should lead to fewer complications during surgery.
Wallwiener and colleagues developed a surgical trainer that, when used in conjunction with
other improvements in their training program, led to lower rates of hysteroscopic
complications. 39,40 However, for most simulators the link between improvements in technical
skills and dexterity from simulator training and prevention of adverse events has yet to be
established and deserves formal investigation. Further, problem-based surgical simulation (eg,
avoiding inadvertent ligation of the ureter during hysterectomy) may improve patient safety not
only by improving skills, but also by training surgeons to better anticipate and avoid
complications and to manage them should they occur.
Simulators in Gastroenterology
Simulators have been developed to train physicians in the technical skills required in
endoscopy. 41-43 We identified one study that evaluated the effect of simulator training on
physician performance. In a small randomized controlled trial enrolling 10 residents, Tuggy and
colleagues found residents who trained for flexible sigmoidoscopy using a virtual reality
simulator were faster, visualized a greater portion of the colon, and made fewer directional errors
in actual patients. 44
Simulators in Cardiology
Cardiology training has long used a variety of simulators from audiocassettes of heart
tones to full patient simulators. Two simulators have been evaluated. Champagne and colleagues
demonstrated that a heart sound simulator could increase medical students’ recognition of
pathologic heart sounds. 45 Ewy and colleagues studied the efficacy of “Harvey,” a cardiology
patient simulator. 46 In a study enrolling 208 senior medical students at 5 medical schools,
participants were randomized to receive training on Harvey versus a standard cardiology
curriculum during their cardiology elective. Students who had been trained with Harvey
performed skills better both when tested with the simulator and when tested with real patients.
Some physicians have expressed concern that training on simulators may decrease
professionalism. In this study, there was no difference in the way patients perceived the
professionalism of the students trained on Harvey compared with students who received standard
training.
Systems that simulate the cardiovascular anatomy and physiology have also been
developed. Swanson and colleagues created a cardiovascular simulator to train physicians on the
workings of mechanical valves and balloon assist devices, and to recognized diseased vessels. 47
As cardiac procedures have become more invasive, simulators to train cardiologist in these
procedures have become more common, 48, 49 but their effect on resident performance has not
been evaluated formally.
Comment
Although simulators have been used for many years in a variety of settings, data on their
efficacy are still emerging. While there is currently no evidence that simulation-based training
leads to improved patient outcome, it may prove difficult to conduct such studies. These would
require large cohorts of patients to be followed during and after care by clinicians who were
randomized to have undergone different cumulative amounts of simulation training. Because
adverse events are uncommon, and there are a large number of patient-based and system-based
514

factors that contribute to negative outcomes, any such study would have to be massive and
prolonged. Instead, provider performance, with its known limitations, has been and will continue
to be used as a surrogate outcome. Nonetheless, as Gaba has asserted “…no industry in which
human lives depend on skilled performance has waited for unequivocal proof of the benefits of
simulation before embracing it.” 50 Certain benefits are clear. In training for procedures,
simulators have high face validity because they ease trainees’ transition to actual patients, which
seems inherently beneficial as a means to avoid adverse events. Further, procedural success is
related to the experience of the operator, known as the volume-outcome relationship 51, 52 (see
Chapter 19). As simulators become more advanced, they may be reasonable substitutes to
improve proficiency of both trainees and low volume physician operators. This increase in
proficiency may have an important impact in patient outcomes. Future studies of the link
between simulator-based training and performance on actual patients will improve our ability to
better assess the appropriate role of simulators in training and patient safety.
The costs of simulators vary widely and need to be considered. “Home-made” or simple
trainers are far less expensive than complex simulators or full-scale simulation centers. The
average cost of high-fidelity patient simulators is on the order of $200,000. Medium fidelity
simulators may be as little as $25,000. Establishing a dedicated simulation center can cost up to
$1,000,000 (including the simulator) depending on the amount of space, the type of clinical
equipment to be used, the extent of renovations needed, and the sophistication of the audiovisual
equipment. However, such capital costs are amortized over a long period of time, and such
centers typically are used for a wide variety of training curricula for diverse target populations.
Further, for most simulation training the dominant cost is that of instructor time. Another indirect
cost is that of removing clinical personnel from revenue producing work to undergo training.
The health care industry currently does not fully embed time or costs of training into the system,
but instead often leaves these costs for the individual clinicians to bear.
There are potential risks to simulation-based training. Where the simulator cannot
properly replicate the tasks or task environment of caring for patients, there is a risk that
clinicians might acquire inappropriate behaviors (negative training) or develop a false sense of
security in their skills that could theoretically lead to harm. Although there are no data to suggest
that this currently happens, such risks will have to be weighed and evaluated as simulators
become more commonly used.
In summary, although there is currently little evidence that simulation training improves
patient care, the experience with simulation in other industries and the high face validity of their
applications in health care has led many institutions to adopt the technology. It is likely that
simulators will continue to be used and their role in training of medical personnel will grow.
Definitive experiments to improve our understanding of their effects on training will allow them
to be used more intelligently to improve provider performance, reduce errors and ultimately,
promote patient safety. Although such experiments will be difficult and costly, they may be
justified to determine how this technology can best be applied.
References
1. Bushell E, Gaba, D.M. Anesthesia Simulation and Patient Safety. Problems in Anesthesia.
2001;In Press.
2. Peugnet F, Dubois P, Rouland JF. Virtual reality versus conventional training in retinal
photocoagulation: a first clinical assessment. Comput Aided Surg. 1998;3:20-26.
3. Gaba DM, DeAnda A. A comprehensive anesthesia simulation environment: re-creating
the operating room for research and training. Anesthesiology. 1988;69:387-394.
515

4. Dunn D. Malignant hyperthermia. AORN J. 1997;65:728,731,passim.
5. Cooper JB, Newbower RS, Long CD, McPeek B. Preventable anesthesia mishaps: a study
of human factors. Anesthesiology. 1978;49:399-406.
6. Spencer FC. Human error in hospitals and industrial accidents: current concepts. J Am Coll
Surg. 2000;191:410-418.
7. Keyser EJ, Derossis AM, Antoniuk M, Sigman HH, Fried GM. A simplified simulator for
the training and evaluation of laparoscopic skills. Surg Endosc. 2000;14:149-153.
8. Gaba DM, DeAnda A. The response of anesthesia trainees to simulated critical incidents.
Anesth Analg. 1989;68:444-451.
9. Gaba DM. Anaesthesia simulators. Can J Anaesth. 1995;42(10):952-953.
10. Gaba DM, Howard SK, Flanagan B, Smith BE, Fish KJ, Botney R. Assessment of clinical
performance during simulated crises using both technical and behavioral ratings.
Anesthesiology. 1998;89:8-18.
11. Ruppert M, Reith MW, Widmann JH, et al. Checking for breathing: evaluation of the
diagnostic capability of emergency medical services personnel, physicians, medical
students, and medical laypersons. Ann Emerg Med. 1999;34:720-729.
12. Dorfsman ML, Menegazzi JJ, Wadas RJ, Auble TE. Two-thumb vs. two-finger chest
compression in an infant model of prolonged cardiopulmonary resuscitation. Acad Emerg
Med. 2000;7(10):1077-1082.
13. Cheung V, Critchley LA, Hazlett C, Wong EL, Oh TE. A survey of undergraduate teaching
in anaesthesia. Anaesthesia. 1999;54:4-12.
14. King PH, Blanks ST, Rummel DM, Patterson D. Simulator training in anesthesiology: an
answer? Biomed Instrum Technol. 1996;30:341-345.
15. DeAnda A, Gaba DM. Unplanned incidents during comprehensive anesthesia simulation.
Anesth Analg. 1990;71:77-82.
16. Schwid HA, O'Donnell D. Anesthesiologists' management of simulated critical incidents.
Anesthesiology. 1992;76:495-501.
17. Schwid HA, Rooke GA, Michalowski P, Ross BK. Screen-based anesthesia simulation
with debriefing improves performance in a mannequin-based anesthesia simulator. Teach
Learn Med. 2001;13(2):92-96.
18. Chopra V, Gesink BJ, de Jong J, Bovill JG, Spierdijk J, Brand R. Does training on an
anaesthesia simulator lead to improvement in performance? Br J Anaesth. 1994;73:293-
297.
19. Gaba DM, Howard,S.K., Fish,K.J., Smith,B.E., Sowb,Y.A. Simulation-based training in
anesthesia crisis resource management (ACRM): a decade of experience. Simulation and
Gaming. 2001;32(2):175-193.
20. Sayre MR, Sakles JC, Mistler AF, Evans JL, Kramer AT, Pancioli AM. Field trial of
endotracheal intubation by basic EMTs. Ann Emerg Med. 1998;31:228-233.
21. New Technologies in anesthesia practice: anesthesia simulators [Audiocassette]. Kansas
City, MO: Audio-Digest Anesthesiology; 1998.
22. Chait P. Future directions in interventional pediatric radiology. Pediatr Clin North Am.
1997;44(3):763-782.
23. Sadler DJ, Parrish F, Coulthard A. Intravenous contrast media reactions: how do
radiologists react? Clin Radiol. 1994;49:879-882.
24. Medina LS, Racadio JM, Schwid HA. Computers in radiology. The sedation, analgesia,
and contrast media computerized simulator: a new approach to train and evaluate
radiologists' responses to critical incidents. Pediatr Radiol. 2000;30:299-305.
516

25. Sica GT, Barron DM, Blum R, Frenna TH, Raemer DB. Computerized realistic simulation:
a teaching module for crisis management in radiology. AJR Am J Roentgenol.
1999;172(2):301-304.
26. Carrico CJ, Satava,R.M. Advanced simulation technologies for surgical education. Bull Am
Coll Surg. 1996;81:77.
27. Ota D, Loftin B, Saito T, Lea R, Keller J. Virtual reality in surgical education. Comput Biol
Med. 1995;25:127-137.
28. Chong CK, How TV, Black RA, Shortland AP, Harris PL. Development of a simulator for
endovascular repair of abdominal aortic aneurysms. Ann Biomed Eng. 1998;26:798-802.
29. Rudman DT, Stredney D, Sessanna D, et al. Functional endoscopic sinus surgery training
simulator. Laryngoscope. 1998;108(11 Pt 1):1643-1647.
30. Ecke U, Klimek L, Muller W, Ziegler R, Mann W. Virtual reality: preparation and
execution of sinus surgery. Comput Aided Surg. 1998;3:45-50.
31. Szekely G, Bajka M, Brechbuhler C, et al. Virtual reality based surgery simulation for
endoscopic gynaecology. Stud Health Technol Inform. 1999;62:351-357.
32. Poss R, Mabrey JD, Gillogly SD, et al. Development of a virtual reality arthroscopic knee
simulator. J Bone Joint Surg Am. 2000;82-A(10):1495-1499.
33. Ballaro A, Briggs T, Garcia-Montes F, MacDonald D, Emberton M, Mundy AR. A
computer generated interactive transurethral prostatic resection simulator. J Urol.
1999;162:1633-1635.
34. Maher JE, Kleinman GE, Lile W, Tolaymat L, Steele D, Bernard J. The construction and
utility of an amniocentesis trainer. Am J Obstet Gynecol. 1998;179:1225-1227.
35. Bettega G, Payan Y, Mollard B, Boyer A, Raphael B, Lavallee S. A simulator for
maxillofacial surgery integrating 3D cephalometry and orthodontia. Comput Aided Surg.
2000;5:156-165.
36. Derossis AM, Bothwell J, Sigman HH, Fried GM. The effect of practice on performance in
a laparoscopic simulator. Surg Endosc. 1998;12:1117-1120.
37. Fried GM, Derossis AM, Bothwell J, Sigman HH. Comparison of laparoscopic
performance in vivo with performance measured in a laparoscopic simulator. Surg Endosc.
1999;13:1077-1081; discussion 1082.
38. Scott DJ, Bergen PC, Rege RV, et al. Laparoscopic training on bench models: better and
more cost effective than operating room experience? J Am Coll Surg. 2000;191:272-283.
39. Wallwiener D, Rimbach S, Bastert G. The HysteroTrainer, a simulator for diagnostic and
operative hysteroscopy. J Am Assoc Gynecol Laparosc. 1994;2:61-63.
40. Wallwiener D, Rimbach S, Aydeniz B, Pollmann D, Bastert G. Operative Hysteroscopy:
Results, Security Aspects, In Vitro Simulation Training (Hysterotrainer). J Am Assoc
Gynecol Laparosc. 1994;1(4, Part 2):S39.
41. Williams CB, Saunders BP, Bladen JS. Development of colonoscopy teaching simulation.
Endoscopy. 2000;32:901-905.
42. Aabakken L, Adamsen S, Kruse A. Performance of a colonoscopy simulator: experience
from a hands-on endoscopy course. Endoscopy. 2000;32:911-913.
43. Neumann M, Mayer G, Ell C, et al. The Erlangen Endo-Trainer: life-like simulation for
diagnostic and interventional endoscopic retrograde cholangiography. Endoscopy.
2000;32(11):906-910.
44. Tuggy ML. Virtual reality flexible sigmoidoscopy simulator training: impact on resident
performance. J Am Board Fam Pract. 1998;11:426-433.
517

45. Champagne MT, Harrell JS, Friedman BJ. Use of a heart sound simulator in teaching
cardiac auscultation. Focus Crit Care. 1989;16:448-456.
46. Ewy GA, Felner JM, Juul D, Mayer JW, Sajid AW, Waugh RA. Test of a cardiology
patient simulator with students in fourth-year electives. J Med Educ. 1987;62:738-743.
47. Swanson WM, Clark RE. A simple cardiovascular system simulator: design and
performance. J Bioeng. 1977;1:135-145.
48. Wang Y, Chui C, Lim H, Cai Y, Mak K. Real-time interactive simulator for percutaneous
coronary revascularization procedures. Comput Aided Surg. 1998;3:211-227.
49. Dawson SL, Cotin S, Meglan D, Shaffer DW, Ferrell MA. Designing a computer-based
simulator for interventional cardiology training. Catheter Cardiovasc Interv. 2000;51:522-
527.
50. Gaba DM. Improving anesthesiologists' performance by simulating reality. Anesthesiology.
1992;76:491-494.
51. Jollis JG, Romano PS. Volume-outcome relationship in acute myocardial infarction: the
balloon and the needle. JAMA. 2000;284:3169-3171.
52. Jollis JG, Peterson ED, Nelson CL, et al. Relationship between physician and hospital
coronary angioplasty volume and outcome in elderly patients. Circulation. 1997;95:2485-
2491.
518

Chapter 46. Fatigue, Sleepiness, and Medical Errors
Ashish K. Jha, MD
University of California, San Francisco School of Medicine
Bradford W. Duncan, MD
Stanford University School of Medicine
David W. Bates, MD, MSc
Harvard Medical School
Introduction
Fatigue may contribute to the human error component of medical errors. 1-3 Hospitals
function around the clock, which necessitates shift work for many personnel. Physicians,
especially those in training, typically work long hours and are often sleep deprived. 4 Personnel
who work during evenings and at night experience disruptions in circadian rhythms, which may
aggravate fatigue. Although little research has focused specifically on fatigue in hospital
personnel and its relationship to medical error, studies outside the medical field demonstrate the
intuitive link between fatigue and degradation in performance and suggest some safety practices
that may be adopted in medicine. Although both acute and chronic fatigue may have detrimental
effects on the health of medical practitioners, 5-7 this chapter focuses on fatigue’s direct effects on
patient safety. We review the literature on problem sleepiness among medical personnel, its
impact on performance, and interventions to address sleep deprivation: limiting work hours,
changes in shift scheduling, napping, and pharmaceutical aids. Although beyond the scope of
this chapter, factors that contribute to fatigue beyond sleepiness, such as job stress and work
load, should be considered as part of a multifaceted strategy to combat fatigue.
Background
Fatigue and sleepiness may affect patient safety in several ways. Physicians and nurses
need good attention, sound judgment, and often quick reaction time, especially in emergency
situations. Whether evaluating an electrocardiogram for signs of myocardial ischemia or
monitoring a patient during general anesthesia, degradation of attention, memory, or
coordination may affect performance and lead to adverse events. Research suggests that sleep
requirements and patterns are idiosyncratic, with wide variation across populations. In order to
design interventions that will effectively decrease or prevent these events, it is important to
understand the signs, prevalence, and impact of sleep deprivation and problem sleepiness.
Sleep Deprivation
Individuals differ in their optimal sleep requirements. Most sleep experts agree that
adults typically need between 6 and 10 hours of sleep per 24-hour period, with most people
requiring approximately 8 hours of sleep per day. 8,9 When adults get less than 5 hours of sleep
over a 24-hour period, peak mental abilities begin to decline. 2 For short periods of time (2-3
days), adult who get 4 hours of sleep can function reasonably well, but below peak levels. 2
However, even with sleep deprivation of just a couple of days, slower response times and
decreased initiatives are observed. 10 After one night of missed sleep, cognitive performance may
decrease 25% from baseline. 11,12 After the second night of missed sleep, cognitive performance
can fall to nearly 40% of baseline. 12
With ongoing sleep deprivation (getting 2 to 3 hours less sleep than optimal), people
develop a sleep debt. 2 If the sleep debt continues over 5 to 10 days, they are rarely maximally
519

alert and at some point general performance, and particularly cognitive performance, become
verifiably worse. Sleep debt also leads to slower response times, altered mood and motivation,
and reduced morale and initiative. A meta-analysis of the effect of sleep deprivation on
performance by Pilcher et al found that humans who are chronically sleep deprived function at
the 9 th percentile of non-sleep-deprived subjects. Further, sleep deprivation affected mood more
than it did cognitive function; both were more affected than motor function. 9
Night-Shifts and Shift Rotation
Shift work usually refers to a schedule in which some employees begin work at times
other than the morning. In hospitals, up to 35% of nurses may be required to work at times other
than the day shift. 13 A report by the Association of Professional Sleep Societies concluded that
night-time operators’ fatigue contributed to 4 well known disasters: Exxon Valdez, Bhopal,
Chernobyl, and Three Mile Island. 14 Fatigue has also been implicated in aircraft accidents 15 and
in poor driving and accidents among truck drivers. 16 It is well documented that shift workers
have disturbances in their circadian rhythm, as measured by changes in their melatonin and
cortisol levels. 17 Sleep after night work tends to be shorter than sleep after day work, leading to
greater cumulative sleep deprivation. 18-20 Shift workers have poorer quality of sleep, marked by
less REM sleep, and are less likely to feel refreshed after awaking. Between 60 and 70 percent of
shift workers complain of sleeping difficulties or problem sleepiness. 21 Several surveys of shift
workers have found that those who work during night shifts are more likely to report sleepiness
at work. 18,19,22,23 Alertness on the job is also affected, with employees showing less alertness
during nighttime shifts. 24 In addition, shift workers tend to perform less well on reasoning and
non-stimulating tasks than non-shift workers. 22,23
Prevalence and Severity
Fatigue and sleep deprivation are common among medical personnel. Long work-hours
are a tradition during residency, 25 with most interns and residents working 80 to 100 hours a
week, often 36 hours at a time. 26 During these shifts their sleep is limited, and is usually
interrupted. 27 In a 1991 national survey, second-year residents reported an average of 37.6 hours
as the largest number of hours without sleep during their first postgraduate year and roughly
25% of the residents reported being on call in the hospital over 80 hours per week. 26 A
movement in the late 1980s, prompted partly by the death of a young woman, 28 led to regulations
in New York State dictating that residents could work a maximum of 80 hours per week, with a
maximum of 24 consecutive hours of patient care, and a minimum of 8 hours off duty between
shifts (see also chapter 55). 29 Despite these regulations, unannounced inspections of 12 teaching
hospitals in New York State in March 1998 found 37% of all residents worked more than 85
hours per week, 20% of all residents and 60% of surgical residents worked more than 95 hours
per week, and 38% of all residents and 67% of all surgical residents worked more than 24
consecutive hours. 30 In 2000, 8% of programs and institutions reviewed by the Accreditation
Council for Graduate Medical Education were cited as being in violation of their work-hour
requirements. 31 Work-hour violations were noted in general surgery (35%), pediatrics (16%),
internal medicine (10%) and other training programs as well. 31
Long hours and sleep deprivation continue after residency. Health care providers,
particularly those still in training or who have recently completed training, occasionally work
extra shifts to increase their income (“moonlighting”). One recent survey found that nearly half
of all emergency medicine residents moonlight. 32 As many as 65% of internal medicine residents
and fellows moonlight 33 and moonlighting is common among other residencies and
520

fellowships. 34, 35 These shifts are often at odd hours, and therefore are disruptive to normal sleep
patterns. Among surgical staff, fatigue is common, especially since surgical teams can be
involved in long, complicated operative cases that can take 12 to 20 hours at a time. 36,37
Multiple studies have documented the impact of fatigue on medical personnel
performance. 38 However, these studies have been limited by poor study designs or outcomes that
may not correlate well with medical error. One study of nursing fatigue suggests that it may play
a role in increased error. Gold and colleagues administered a questionnaire to nurses at a large
academic hospital and found that nurses who worked a rotating schedule, when compared with
nurses who predominantly worked day shifts, were more likely to fall asleep at work and get less
sleep over all, and were nearly twice as likely to report committing a medication error. 39
Using standardized testing, investigators have found that after a night of call, sleep
deprived physicians may have worse language and numeric skills, 40 retention of information, 41
short-term memory, 42 and concentration. 43 Performance on standardized tests may not reflect
performance in medical situations. Taffinder et al studied the impact of sleep deprivation on
surgical residents previously trained on a simulator and found that after a night without sleep,
surgeons were slower and more prone to errors on the simulator than those who had a normal
night of sleep. 44 Similarly, Denisco et al studied anesthesia residents after a night of sleep
deprivation and found that those who had been on call and were sleep deprived scored less well
on simulated critical events. 45 Smith-Coggins et al compared cognitive and motor performance
of emergency physicians and found that, as the 24-hour study period progressed, physicians were
more likely to make errors during a simulated triage test and while intubating a mannequin. 19
However, other studies have failed to find an effect of sleep deprivation on cognitive
performance by resident physicians. 46-48 Simulators may not reflect actual medical performance
(see chapter 45). Though psychomotor performance seems to be affected by sleep deprivation,
data are inconsistent as to fatigue’s impact on cognitive function and there are inadequate data
assessing its impact on clinical performance.
Few studies have looked at the impact of fatigue in hospital personnel on adverse events.
A retrospective study by Haynes et al of 6371 surgical cases, found that the risk of postoperative
complications among patients undergoing surgery was not increased when the surgical resident
was sleep deprived. 49 These results may not be surprising for several reasons. First, the authors
did not measure the residents’ error rate, which may have been higher with sleep deprivation.
Second, the study did not measure the role attending physicians or other operating room
personnel may have played in averting adverse events when residents erred. The supervisory
aspect of system design can (and should) reduce both the frequency of individual mistakes (error
prevention) and the likelihood of adverse events given that errors are inevitable (error
absorption). 1 Finally, the rate of adverse events, including those that did not result in operative
complications (“near misses”), may have been higher but under reported. Well-designed studies
that evaluate the effects of fatigue among medical personnel on rates of medical errors or adverse
events would be useful. In the meantime, the lack of convincing data linking fatigue with poor
patient outcomes should not deter us from tackling the issue of fatigue among medical personnel.
Practice Descriptions
Hours of Service
We reviewed the evidence for 2 potential safety practices concerning hours of service: 8hour
versus 12-hour length shifts and regulations limiting maximum shift length and/or total
hours worked. Most observational studies on optimal shift length to reduce fatigue and maximize
521

performance are in non-medical settings and present inconsistent findings. In a study of workplace
accidents in Germany, Hanecke et al found accident risk increased exponentially after the
9 th hour at work and was highest among workers whose shift began in the evening or night. 50 The
authors concluded that shifts that last longer than 8 hours might lead to more worker fatigue and
higher risk of accidents. Axelsson and colleagues studied workers at a power plant and found no
difference in sleepiness or performance between those who worked 8-hour shifts and those who
worked 12-hour shifts. 51 Another group found that switching from 8- to 12-hour shifts led to
increased alertness on the job and improved recovery time after night shifts. 52 Overland has
proposed that work that requires complex cognitive tasks may be ill suited for longer shifts,
whereas work with limited cognitive demands may be well suited for longer shifts. 53 Because the
components of work vary dramatically within and across industries, shift durations that maintain
performance in one setting may be ineffective in another.
We identified 9 observational studies comparing 8- versus 12-hour shifts for medical
personnel. Two studies of nursing care on 10 wards found that quantity 54 and quality 55 of care
were significantly lower with 12-hour shifts. Six studies of nurses 56-61 and one of physicians 62
measured outcomes including self-reported alertness, self-reported performance, and/or worker
satisfaction. While 2 nurse studies found that self-reported alertness, performance, and
satisfaction wane with longer shifts, 56,57 Urgovics and Wright found that ICU nurses reported
higher job satisfaction and subjectively improved clinical performance with 12-hour shifts. 60 The
3 remaining studies in nurses found no difference in either satisfaction or self-reported
performance between 8- and 12-hour shifts. 58,59,61 A survey of emergency department physicians
found that those who worked 12-hour shifts were less likely to be satisfied than those who
worked 8-hour shifts. 62 The relationship between these subjective outcomes measures and
medical error is not clear.
Hours of service regulations as an effort to reduce errors due to fatigue are standard in
some non-medical fields. Truck drivers are typically allowed to work no more than 10 hours at a
time and no more than 60 hours in one week. Airline pilots and air traffic controllers work
regulated hours and some data suggest waning performance as work-hours increase. 24,63-65
Although most health care personnel are not subject to work-hour standards, many physiciansin-training
are, either by statutory regulations or by being in an accredited training program. In a
retrospective cohort study, Laine and colleagues found the aforementioned New York State
regulations limiting resident work-hours had no effect on patient outcomes such as mortality or
transfers to the intensive care unit but were associated with increased rates of medical
complications and delays in diagnostic tests. 66 These negative effects may have been related to
discontinuity of care and/or fewer physician-hours per patient. As the authors noted, “better care
may be provided by a tired physician who is familiar with the patient than by a rested physician
who is less familiar with the patient.” 66 In a case-control study, Petersen and colleagues found
that when patients were cared for by a physician other than their primary resident, they were 6
times as likely to suffer a preventable adverse event. 67 Thus, fewer physician work hours may
lead to more physician discontinuity and potentially, more adverse events and poorer outcomes
for patients.
On the other hand, Gottlieb studied changes in a medical service staffing schedule that
allowed for reduced sleep deprivation, improved distribution of admissions throughout the week,
and improved continuity of inpatient care. 68 After these changes were instituted, patients had
shorter lengths of stay, fewer ancillary tests, and fewer medication errors. Although it is difficult
to ascribe the improvements to changes in work-hours because several other changes were made
as well, it does appear that changes in work-hours can be made without adversely affecting
522

patient outcomes. Any effort to change duty hours for health care personnel in an effort to reduce
fatigue should factor in and continuously monitor numerous variables, including the potential
costs of discontinuity, medical complications and unnecessary hospital days, to ensure that the
measures do not compromise patient care. The costs needed to maintain adequate staffing in face
of lost physician work-hours has been estimated to be $360 million in New York State alone. 69
However, the difficult task of estimating other costs and potential savings from implementing
these regulations has not been accomplished.
Finally, some authors have expressed concern that restriction of resident physician workhours
may lead to poorer quality training and decreased professionalism among doctors. 70 They
argue that restricted working hours will decrease a sense of obligation to patients and will
sanction self-interest over the well-being of patients. However, there are no data to substantiate
these concerns.
Direction and Speed of Rotation of Shift Work
The direction of shift rotation may impact worker fatigue. For workers who change from
one shift to another, a forward rotation of shift work (morning shifts followed by evening shifts
followed by night shifts) may lead to less fatigue on the job than backward rotation (day shift to
night shift to evening shift). 71-74 Forward rotation appears easier to tolerate physiologically since
the natural circadian rhythm tends to move forward and it is more difficult to fall asleep earlier
than the normal bedtime. Several studies in non-medical personnel have shown that forward
rotation allows for better acclimation of the circadian rhythm. 2,12,75 However, 2 other studies
found no significant difference in forward versus backward shift rotation. 76,77 None of these
studies measured worker performance or error rates and we found no studies that evaluated
direction of shift work rotation among medical personnel.
Another variable in scheduling is the speed of shift work rotation. Studies suggest that
slow rotation (eg, changing from one shift to another every one to two weeks) may allow for
better adaptation of the circadian rhythm than fast rotation (eg, changing shifts every 2-3
days). 71,73,78,79 Slow shift rotation results in greater sleep length at home, less sleepiness on the
job, better self-reported performance, and fewer errors. 74,79 In some cases, fast rotation may
increase worker satisfaction 80 but the effects of such satisfaction on safety have not been
assessed. Shift rotation at an extremely slow rate approximates fixed, non-rotating shifts
(permanent night shifts, permanent day shifts). Permanent shifts are associated with better
adaptation to changes in the circadian rhythm 78 and better performance than rotating shifts. 79
However, daytime commitments and social obligations often prevent workers from completely
adapting to permanent night shifts and worker satisfaction is poor. 71
Improving Sleep: Education about Sleep Hygiene
Good sleep hygiene, including the avoidance of alcohol and caffeine before bedtime, and
maintaining a healthy sleep environment, may aid in decreasing sleep debt and fatigue. Studies
of sleep hygiene have focused on treatment of persons with insomnia or other chronic sleep
disorders. 81-83 We found no clinical studies that measure the efficacy of good sleep hygiene
among shift workers. Generally, most employers cannot dictate how their workers spend their
hours off-duty and compliance with recommendations may be poor. One study of lawenforcement
officers working rotating shifts found significant increases in awareness and
knowledge after a training session on sleep hygiene practices but no change on a post-sleep
523

inventory assessed at one-month follow-up. 84 The effectiveness of educational programs about
sleep hygiene to improve shift worker performance requires further study.
Lighting at Work
The body’s regulation of circadian rhythm is mediated by the effects of light and
darkness. A 1986 survey found that 7.3 million Americans work at night. 71 These employees,
who work during dark hours and sleep during daylight hours, are often chronically sleep
deprived and may suffer adverse health effects, 85 partially due to poor synchrony of circadian
rhythm to work schedule. Since scheduled light exposure can produce a phase shift in the
endogenous circadian rhythm, 71, 86 investigators have studied changes in lighting at work and
home to improve adjustment to the shift cycle. Foret et al studied 8 young men in a sleep lab and
found exposure to bright lights during the night produced a beneficial effect on subjective
alertness. 87 Czeisler and colleagues found that subjects who were exposed to bright light at night
and nearly complete darkness during the day had better cognitive performance and subjective
alertness, and longer daytime sleep (7.7 vs. 5.7 hours, p=0.01). 88
Manipulation of light and dark is much easier in sleep labs than in the field, 89 where
unintended exposure to bright light is common and may adversely impact attempts to alter
workers’ circadian rhythm. 90 The National Aeronautics and Space Administration (NASA) has
studied the efficacy of bright lights on shuttle astronauts. Their encouraging results suggest that
alterations in circadian rhythm can be obtained upon on exposure to light at night. 89,91 The
United States Nuclear Regulatory Commission has also implemented bright lighting for its night
workers and found less fatigue and better alertness on the job. 92 Field studies are needed to
determine how bright artificial light affects objective measures of performance in health care
workers and medical error. Bright light may not be appropriate for all areas of the hospital. For
example, Bullough and Rea have noted that while bright light might help workers in neonatal
care units, it may also be detrimental to patients. 93
Nonetheless, lighting can be a relatively inexpensive intervention using existing
equipment. Keeping lights bright at night, and educating workers about using heavy shades at
home may have an important impact on worker performance on night shifts.
Napping
Napping is common among shift workers and is perceived as a way to combat
fatigue. 94,95 One study of shift workers in a steel plant found that over half reported napping at
home either before or after their shifts. 94 The efficacy of naps has been studied in 3 settings:
prior to periods of sleep deprivation (prophylactic naps), during periods of sleep deprivation
(therapeutic naps) and during work hours (maintenance naps). Most studies have been
conducted in sleep labs in healthy, young, male subjects.
A number of studies in the non-medical literature have studied the efficacy of
prophylactic napping. Gillberg and colleagues studied 8 male subjects who were allowed only 4
hours of sleep at night. When subjects took a 30-minute nap in the middle of the prior day, they
had better subjective alertness, 20% improvement in vigilance performance, and less overall
sleepiness than when they had not been allowed to nap. 96 Others have also found benefits of
prophylactic naps on subjective and objective measures of alertness and performance in healthy
volunteers undergoing extended periods of sleep deprivation. 97-100 Bonnet and Arand studied
prophylactic versus therapeutic naps in 12 healthy young men who underwent 24 hours of sleep
deprivation to simulate sleep patterns of medical housestaff. 101 One group of subjects had a 4-
524

hour prophylactic nap in the evening and caffeine during the 24 hours, while the second group
had four, 1-hour naps during the 24-hour work period and no caffeine. Those in the prophylactic
nap and caffeine group had a 15% increase in reasoning and overall improved subjective
alertness compared with the group that had only short naps. There was no impact on mood. We
identified one study of napping by medical personnel. Harma and colleagues studied 146 female
hospital nurses and nurses’ aides and found that those who napped prior to their night shifts were
less likely to report on the job fatigue. 95
Most studies evaluating the efficacy of therapeutic napping during prolonged periods of
sleep deprivation have found beneficial effects when compared with no napping. 102-108 On the
other hand, Gillberg and colleagues found no difference in simulated driving between the 2
groups of sleep deprived truck drivers, one group having taken a 30-minute nap during the
middle of the previous night. 109
Maintenance naps are naps that occur on the job, during the shift. These naps could
compensate for daytime sleep deprivation or could bridge the nighttime low point in circadian
somnolence. 110 Many Japanese industries have provided their employees with the option of on
the job napping and nearly half of nighttime shift workers take advantage of this opportunity. 111
Though no systematic studies of the impact of maintenance naps exist in shift workers, one
investigation found that short naps in the middle of the night improved performance for the rest
of the shift. 98 Napping over several successive shifts has not been studied. 110
An important consideration in napping is the phenomena of sleep inertia, a period of
transitory hypo-vigilance, confusion, disorientation of behavior and impaired cognitive
performance that immediately follows awakening. 112 Sleep inertia is well documented 112-116 and
lasts up to 30 minutes after awakening. 116-118 The duration of deep sleep and the time of the nap,
relative to the circadian cycle, seem most related to the severity of sleep inertia. 8 Strategies for
napping on the job to reduce fatigue should be designed to avoid possible detrimental effects of
sleep inertia. Another potential negative effect of lengthy naps is that they can disrupt the
quantity and quality of later sleep periods. 119
In summary, there is strong evidence that therapeutic naps and maintenance naps combat
the effects of fatigue and sleep loss. They can help subjects adapt better to circadian rhythm
disturbances and perform better during acute sleep deprivation. Their application in the medical
field is not well known. While prophylactic and therapeutic napping result in loss of social time
at home, maintenance napping results in loss of work time. Costs associated with naps have not
been reported. The financial impact of reduced worker fatigue due to napping has not been
evaluated in medicine.
Medical Therapies
Melatonin is the major hormone responsible for circadian rhythm regulation. James et al
studied the effect of oral melatonin supplementation on circadian rhythm and adaptation to night
shifts among medical personnel. 120 They and others have found no effect among medical shift
workers. 121-123 Though melatonin continues to be studied for chronic insomnia and other
conditions, there currently is insufficient evidence to recommend its use to combat the fatigue
associated with changing workshifts.
Some studies have looked at the potential benefits of benzodiazepines and other sedatives
for short-term insomnia associated with shift work, but no data exist on long-term use.
Stimulants and caffeine can boost performance acutely but do not address the underlying sleep
deprivation, 124 and thus are not a viable long-term solution. Furthermore, concern over side
525

effects, addiction, and performance degradation with current pharmacologic interventions makes
their use as a safety practice unlikely.
Comment
Sleep deprivation and disturbances of circadian rhythm lead to fatigue, decreased
alertness, and poor performance on standardized testing. Although data from non-medical fields
suggest that sleep deprivation leads to poor job performance, this link has not yet been
established in medicine. Although the link with fatigue seems intuitive, promoting interventions
designed to combat medical errors should be evidence-based. Limits on physician duty hours
must account for potentially detrimental effects of discontinuity in patient care. Forward rather
than backward shift rotation, education about good sleep hygiene, and strategic napping before
or during shifts may reduce fatigue and improve performance. High face validity, low likelihood
of harm, and ease of implementation make these promising strategies, although more evidence of
their effectiveness in medicine is warranted. Studies on the use of bright light in the medical
workplace are needed before it can be embraced.
As Gaba points out, 125 in most high-hazard industries the assumption is that fatigue and
long, aberrant work hours lead to poor performance, and the burden of proof is in the hands of
those who believe that such work practices are safe. In medicine, concerns over discontinuity of
care, and difficulties in changing medical culture have pushed the burden of proof into the hands
of those who wish to change the status quo. Given that medical personnel, like all human beings,
probably function suboptimally when fatigued, efforts to reduce fatigue and sleepiness should be
undertaken, and the burden of proof should be in the hands of the advocates of the current
system to demonstrate that it is safe.
Finally, fatigue among medical personnel may not be fully remediable and human errors
are, in the end, inevitable. The ultimate solution for health care organizations will likely require a
systems-based approach that both limits the potential for human error and intercepts errors that
do occur before they reach patients.
References
1. Leape LL. Error in medicine. JAMA.1994; 272:1851-1857.
2. Krueger GP. Fatigue, Performance, and Medical Error. In: Bogner MS, ed. Human Error in
Medicine. Hillsdale, N.J.: L. Erlbaum Associates; 1994:311, 326.
3. Kohn LT, Corrigan J, Donaldson MS. To err is human: building a safer health system.
Washington, DC: National Academy Press; 2000.
4. Lagnado L. Hospitals overwork young doctors in New York. Wall Street Journal. May 19,
1998, 1998.
5. Marcus CL, Loughlin GM. Effect of sleep deprivation on driving safety in housestaff.
Sleep. 1996;19:763-766.
6. Schwartz AJ, Black ER, Goldstein MG, Jozefowicz RF, Emmings FG. Levels and causes
of stress among residents. J Med Educ. 1987;62:744-753.
7. Small GW. House officer stress syndrome. Psychosomatics. 1981;22:860-869.
8. Rosekind MR, Gander PH, Gregory KB, et al. Managing fatigue in operational settings 1:
Physiological considerations and counter-measures. Hosp Top. 1997;75:23-30.
9. Pilcher JJ, Huffcutt AI. Effects of sleep deprivation on performance: a meta-analysis.
Sleep. 1996;19:318-326.
10. Koslowsky M, Babkoff H. Meta-analysis of the relationship between total sleep
deprivation and performance. Chronobiol Int. 1992;9:132-136.
526

11. Cox T K, GP, ed. Stress and Sustained Performnace. 1 ed; 1989. Work and Stress; No. 3.
12. Krueger G, ed. Sustained work, fatigue, sleep loss and performance: A review of the issues.
1 ed; 1989. Work and Stress; No. 3.
13. Coffey LC, Skipper JK, Jr., Jung FD. Nurses and shift work: effects on job performance
and job-related stress. J Adv Nurs. 1988;13:245-254.
14. Mitler MM, Carskadon MA, Czeisler CA, Dement WC, Dinges DF, Graeber RC.
Catastrophes, sleep, and public policy: consensus report. Sleep. 1988;11:100-109.
15. Rosekind MR, Gregory, K.B., Miller, D.L., Co EL, Lebacqz, J.V. Aircraft Accident
Report: Uncontrolled Collision with Terrain, American International Airways Flight 808,
Douglas DC-8, N814CK, U.S. Naval Air Station, Guantanamo Bay, Cuba, August 18,1993:
National Transportation Safety Board; 1994.
16. Wylie CD, Shultz, T., Miller, J.C. Commercial motor vehicle driver fatigue and alertness
study: Essex Corporation; October 1996.
17. Akerstedt T, Levi L. Circadian rhythms in the secretion of cortisol, adrenaline and
noradrenaline. Eur J Clin Invest. 1978;8:57-58.
18. Salvendy G. Handbook of human factors and ergonomics. 2nd ed. New York: Wiley; 1997.
19. Smith-Coggins R, Rosekind MR, Hurd S, Buccino KR. Relationship of day versus night
sleep to physician performance and mood. Ann Emerg Med. 1994;24:928-934.
20. Smith-Coggins R, Rosekind MR, Buccino KR, Dinges DF, Moser RP. Rotating shiftwork
schedules: can we enhance physician adaptation to night shifts? Acad Emerg Med.
1997;4:951-961.
21. Akerstedt T, Torsvall L. Shift work. Shift-dependent well-being and individual differences.
Ergonomics. 1981;24:265-273.
22. Akerstedt T, Kecklund G, Knutsson A. Manifest sleepiness and the spectral content of the
EEG during shift work. Sleep. 1991;14:221-225.
23. Akerstedt T. Sleepiness as a consequence of shift work. Sleep. 1988;11:17-34.
24. Tucker P, Smith L, Macdonald I, Folkard S. Shift length as a determinant of retrospective
on-shift alertness. Scandinavian Journal of Work, Environment and Health. 1998;24 Suppl
3:49-54.
25. Green MJ. What (if anything) is wrong with residency overwork? Ann Intern Med.
1995;123:512-517.
26. Daugherty SR, Baldwin DC, Jr., Rowley BD. Learning, satisfaction, and mistreatment
during medical internship: a national survey of working conditions. JAMA. 1998;279:1194-
1199.
27. Harvey R, Jarrett PG, Peltekian KM. Patterns of paging medical interns during night calls
at two teaching hospitals. CMAJ. 1994;151:307-311.
28. Asch DA, Parker RM. The Libby Zion case. One step forward or two steps backward? N
Engl J Med. 1988;318:771-775.
29. New York State Health Code. Section 405; 1989.
30. DeBuono BA, Osten WM. The medical resident workload: the case of New York State.
JAMA. 1998;280:1882-1883.
31. Education ACfGM. Percent of programs cited for work hours 1999 and 2000:
Accreditation Council for Graduate Medical Education; 2001 May 9 2001.
32. Li J, Tabor R, Martinez M. Survey of moonlighting practices and work requirements of
emergency medicine residents. Am J Emerg Med. 2000;18:147-151.
33. McCue JD, Janiszewski M, Stickley WT. Residents' views of the value of moonlighting.
Arch Intern Med. 1990;150:1511-1513.
527

34. Majidian AM, Brinker MR, Rice JC, Kerstein MD. Moonlighting in a university surgical
training program. South Med J. 1993;86:441-446.
35. Hunt KR, Hillman BJ, Witzke DB. Moonlighting during the radiology residency. Invest
Radiol. 1992;27:978-983.
36. Chalian AA, Kagan SH. Backside first in head and neck surgery?: preventing pressure
ulcers in extended length surgeries. Head Neck. 2001;23:25-28.
37. Greenberg MS. Handbook of Neurosurgery. Vol 4. Lakeland, Florida: Greenberg Graphics,
Inc.; 1997.
38. Samkoff JS, Jacques CH. A review of studies concerning effects of sleep deprivation and
fatigue on residents' performance. Acad Med. 1991;66:687-693.
39. Gold DR, Rogacz S, Bock N, et al. Rotating shift work, sleep, and accidents related to
sleepiness in hospital nurses. Am J Public Health. 1992; 82:1011-1014.
40. Hawkins MR, Vichick DA, Silsby HD, Kruzich DJ, Butler R. Sleep and nutritional
deprivation and performance of house officers. J Med Educ. 1985;60:530-535.
41. Hart RP, Buchsbaum DG, Wade JB, Hamer RM, Kwentus JA. Effect of sleep deprivation
on first-year residents' response times, memory, and mood. J Med Educ. 1987;62:940-942.
42. Rubin R, Orris P, Lau SL, Hryhorczuk DO, Furner S, Letz R. Neurobehavioral effects of
the on-call experience in housestaff physicians. J Occup Med. 1991;33:13-18.
43. Robbins J, Gottlieb F. Sleep deprivation and cognitive testing in internal medicine house
staff. West J Med. 1990;152:82-86.
44. Taffinder NJ, McManus IC, Gul Y, Russell RC, Darzi A. Effect of sleep deprivation on
surgeons' dexterity on laparoscopy simulator. Lancet. 1998;352:1191.
45. Denisco RA, Drummond JN, Gravenstein JS. The effect of fatigue on the performance of a
simulated anesthetic monitoring task. J Clin Monit. 1987;3:22-24.
46. Reznick RK, Folse JR. Effect of sleep deprivation on the performance of surgical residents.
Am J Surg. 1987;154:520-525.
47. Deaconson TF, O'Hair DP, Levy MF, Lee MB, Schueneman AL, Codon RE. Sleep
deprivation and resident performance. JAMA. 1988;260:1721-1727.
48. Weinger MB, Vora, S. Hendon, C.N., Howard, S.K, Smith, B.E., Mazzei, W.J., Rosekind,
M.R., Gaba D.M. Evaluation of the Effects of Fatigue and Sleepiness on Clinical
Performance in On-call Anesthesia Residents During Actual Nighttime cases and
simulated cases. Paper presented at: Evaluating Patient Safety and Reducing Errors in
Health Care; November 8-10, 1998, 1998; Rancho Mirage, CA.
49. Haynes DF, Schwedler M, Dyslin DC, Rice JC, Kerstein MD. Are postoperative
complications related to resident sleep deprivation? South Med J. 1995; 88:283-289.
50. Hänecke K, Tiedemann S, Nachreiner F, Grzech-Sukalo H. Accident risk as a function of
hour at work and time of day as determined from accident data and exposure models for the
German working population. Scandinavian Journal of Work, Environment and Health.
1998; 24 Suppl 3:43-48.
51. Axelsson J, Kecklund G, Akerstedt T, Lowden A. Effects of alternating 8- and 12-hour
shifts on sleep, sleepiness, physical effort and performance. Scandinavian Journal of Work,
Environment and Health. 1998; Vol 24 Suppl 3:62-68.
52. Lowden A, Kecklund G, Axelsson J, Akerstedt T. Change from an 8-hour shift to a 12hour
shift, attitudes, sleep, sleepiness and performance. Scandinavian Journal of Work,
Environment and Health. 1998; 24 Suppl 3:69-75.
528

53. Overland DW. Comparison of Effects of Change From 8 to 12 hour shifts on Air Force
Aircraft Maintenance Workers. Alabama: Graduate School of Logistics Acquisition
Management, Air Force Institute of Technology; 1997.
54. Reid N, Robinson G, Todd C. The quantity of nursing care on wards working 8- and 12hour
shifts. Int J Nurs Stud. 1993;30:403-413.
55. Todd C, Reid N, Robinson G. The quality of nursing care on wards working eight and
twelve hour shifts: a repeated measures study using the MONITOR index of quality of
care. Int J Nurs Stud. 1989;26:359-368.
56. Smith L, Folkard S, Tucker P, Macdonald I. Work shift duration: a review comparing eight
hour and 12 hour shift systems. Occup Environ Med. 1998;55:217-229.
57. Todd C, Robinson G, Reid N. 12-hour shifts: job satisfaction of nurses. J Nurs Manag.
1993;1:215-220.
58. Fields WL, Loveridge C. Critical thinking and fatigue: how do nurses on 8- & 12-hour
shifts compare? Nurs Econ. 1988;6:189-195.
59. Mills ME, Arnold B, Wood CM. Core-12: a controlled study of the impact of 12-hour
scheduling. Nurs Res. 1983;32:356-361.
60. Ugrovics A, Wright J. 12-hour shifts: does fatigue undermine ICU nursing judgments?
Nurs Manage. 1990;21:64A, 64D, 64F-64G.
61. Washburn MS. Fatigue and critical thinking on eight-and twelve-hour shifts. Nurs Manage.
1991;22:80A, 80D, 80F-80H.
62. Thomas H, Jr., Schwartz E, Whitehead DC. Eight- versus 12-hour shifts: implications for
emergency physicians. Ann Emerg Med. 1994;23:1096-1100.
63. Luna TD. Air traffic controller shiftwork: what are the implications for aviation safety? A
review. Aviat Space Environ Med. 1997;68:69-79.
64. Price WJ, Holley DC. Shiftwork and safety in aviation. Occup Med. 1990;5:343-377.
65. Martin TC. Adverse effects of rotating schedules on the circadian rhythms of air medical
crews. Air Med J. 1995;14:83-86.
66. Laine C, Goldman L, Soukup JR, Hayes JG. The impact of a regulation restricting medical
house staff working hours on the quality of patient care. JAMA. 1993; 269:374-378.
67. Petersen LA, Brennan TA, O'Neil AC, Cook EF, Lee TH. Does housestaff discontinuity of
care increase the risk for preventable adverse events? Ann Intern Med. 1994; 121:866-872.
68. Gottlieb DJ, Parenti CM, Peterson CA, Lofgren RP. Effect of a change in house staff work
schedule on resource utilization and patient care. Arch Intern Med. 1991; 151:2065-2070.
69. Thorpe KE. House staff supervision and working hours. Implications of regulatory change
in New York State. JAMA. 1990;263:3177-3181.
70. Holzman IR, Barnett SH. The Bell Commission: ethical implications for the training of
physicians. Mt Sinai J Med. 2000;67:136-139.
71. Czeisler CA, Moore-Ede MC, Coleman RH. Rotating shift work schedules that disrupt
sleep are improved by applying circadian principles. Science. 1982;217:460-463.
72. Akerstedt T. Shifted sleep hours. Ann Clin Res. 1985;17:273-279.
73. Folkard S. Shiftwork and performance. Paper presented at: Symposium on the Variations
of Work-Sleep Schedules, 1981.
74. Knauth P. Speed and direction of shift rotation. J Sleep Res. 1995; 4:41-46.
75. Comperatore C.A. K, G.P. Circadian rhythm desynchronosis, jet lag, shift lag, and coping
strategies. In: A.J. S, ed. Occupational medicine: Shiftwork. State of the art reviews.
Philadelphia, PA: Hanley & Belfus; 1990:323-341.
529

76. Tucker P, Smith L, Macdonald I, Folkard S. Effects of direction of rotation in continuous
and discontinuous 8 hour shift systems. Occup Environ Med. 2000;57:678-684.
77. Barton J, Folkard S, Smith L, Poole CJ. Effects on health of a change from a delaying to an
advancing shift system. Occup Environ Med. 1994;51:749-755.
78. Knauth P, Rutenfranz J, Schulz H, et al. Experimental shift work studies of permanent
night, and rapidly rotating, shift systems. II. Behaviour of various characteristics of sleep.
Int Arch Occup Environ Health. 1980;46:111-125.
79. Pilcher JJ, Lambert BJ, Huffcutt AI. Differential effects of permanent and rotating shifts on
self-report sleep length: a meta-analytic review. Sleep. 2000;23:155-163.
80. Williamson AM, Sanderson JW. Changing the speed of shift rotation: a field study.
Ergonomics. 1986;29:1085-1095.
81. Morin CM, Colecchi C, Stone J, Sood R, Brink D. Behavioral and pharmacological
therapies for late-life insomnia: a randomized controlled trial. JAMA. 1999;281:991-999.
82. Friedman L, Benson K, Noda A, et al. An actigraphic comparison of sleep restriction and
sleep hygiene treatments for insomnia in older adults. J Geriatr Psychiatry Neurol.
2000;13:17-27.
83. Vgontzas AN, Kales A. Sleep and its disorders. Annu Rev Med. 1999;50:387-400.
84. Holbrook MI, White MH, Hutt MJ. Increasing awareness of sleep hygiene in rotating shift
workers: arming law-enforcement officers against impaired performance. Percept Mot
Skills. 1994;79:520-522.
85. Monk TH, Folkard, S. Making Shift Work Tolerable. London: Taylor and Frances; 1992.
86. Budnick LD, Lerman SE, Nicolich MJ. An evaluation of scheduled bright light and
darkness on rotating shiftworkers: trial and limitations. Am J Ind Med. 1995; 27:771-782.
87. Foret J, Daurat A, Tirilly G. Effect of bright light at night on core temperature, subjective
alertness and performance as a function of exposure time. Scand J Work Environ Health.
1998; 24:115-120.
88. Czeisler CA, Johnson MP, Duffy JF, Brown EN, Ronda JM, Kronauer RE. Exposure to
bright light and darkness to treat physiologic maladaptation to night work. N Engl J Med.
1990;322:1253-1259.
89. Eastman CI, Boulos Z, Terman M, Campbell SS, Dijk DJ, Lewy AJ. Light treatment for
sleep disorders: consensus report. VI. Shift work. J Biol Rhythms. 1995;10:157-164.
90. Chesson AL, Jr., Littner M, Davila D, et al. Practice parameters for the use of light therapy
in the treatment of sleep disorders. Standards of Practice Committee, American Academy
of Sleep Medicine. Sleep. 1999;22:641-660.
91. Putcha L DD, Nimmagudda R, Stewart KT, Kripke DF. Chronobiological measurements in
astronauts. Sleep Research. 1997;26:746.
92. Baker TL. Use of a circadian lighting system to improve night shift alertness and
performance at the USNRC's Headquarters Operations Center. Paper presented at: Safety
of Operating Nuclear Reactions; September 17-20, 1995, 2001; Seattle, WA.
93. Bullough J, Rea, M.S. Lighting for neonatal intensive care units: some critical information
for design. Lighting Research and Technology. 1996;28:189-198.
94. Akerstedt T, Torsvall L. Napping in shift work. Sleep. 1985;8:105-109.
95. Harma M, Knauth P, Ilmarinen J. Daytime napping and its effects on alertness and shortterm
memory performance in shiftworkers. Int Arch Occup Environ Health. 1989;61:341-
345.
96. Gillberg M, Kecklund G, Axelsson J, Akerstedt T. The effects of a short daytime nap after
restricted night sleep. Sleep. 1996;19:570-575.
530

97. Nicholson AN, Pascoe PA, Roehrs T, et al. Sustained performance with short evening and
morning sleeps. Aviat Space Environ Med. 1985;56:105-114.
98. Gillberg M. The effects of two alternative timings of a one-hour nap on early morning
performance. Biol Psychol. 1984;19:45-54.
99. Dinges DF, Orne MT, Whitehouse WG, Orne EC. Temporal placement of a nap for
alertness: contributions of circadian phase and prior wakefulness. Sleep. 1987;10:313-329.
100. Taub JM. Effects of scheduled afternoon naps and bedrest on daytime alertness. Int J
Neurosci. 1982;16:107-127.
101. Bonnet MH, Arand DL. Impact of naps and caffeine on extended nocturnal performance.
Physiol Behav. 1994;56:103-109.
102. Taub JM, Tanguay PE, Clarkson D. Effects of daytime naps on performance and mood in a
college student population. J Abnorm Psychol. 1976;85:210-217.
103. Lubin A, Hord DJ, Tracy ML, Johnson LC. Effects of exercise, bedrest and napping on
performance decrement during 40 hours. Psychophysiology. 1976;13:334-339.
104. Lumley M, Roehrs T, Zorick F, Lamphere J, Roth T. The alerting effects of naps in sleepdeprived
subjects. Psychophysiology. 1986;23:403-408.
105. Mullaney DJ, Kripke DF, Fleck PA, Johnson LC. Sleep loss and nap effects on sustained
continuous performance. Psychophysiology. 1983;20:643-651.
106. Matsumoto K. Effects of nighttime naps on body temperature changes, sleep patterns, and
self-evaluation of sleep. J Hum Ergol (Tokyo). 1981;10:173-184.
107. Naitoh P, Englund CE, Ryman D. Restorative power of naps in designing continuous work
schedules. J Hum Ergol (Tokyo). 1982;11:259-278.
108. Rosekind MR, Graeber, R. C., Dinges, D. F., Connell, L. J., Rountree, M. S., Spinweber C.
L., Gillen, K. A. Crew Factors in Flight Operations IX: Effects of Planned Cockpit Rest on
Crew Performance and Alertness in Long-Haul Operations. Moffett Field, California:
NASA; 1994.
109. Gillberg M, Kecklund G, Akerstedt T. Sleepiness and performance of professional drivers
in a truck simulator- -comparisons between day and night driving. J Sleep Res. 1996;5:12-
15.
110. Rosa RR, Bonnet MH, Bootzin RR, et al. Intervention factors for promoting adjustment to
nightwork and shiftwork. Occup Med. 1990;5:391-415.
111. Reinberg A, Andlauer P, Vieux N, Permanent Commission and International Association
on Occupational Health. Scientific Committee on Shift Work. Night and shift work :
biological and social aspects : proceedings of the Fifth International Symposium on Night
and Shift Work : Scientific Committee on Shift Work of the Permanent Commission and
International Association on Occupational Health (PCIAOH) Rouen, 12-16 May 1980. 1st
ed. Oxford ; New York: Pergamon Press; 1981.
112. Ferrara M, De Gennaro L. The sleep inertia phenomenon during the sleep-wake transition:
theoretical and operational issues. Aviat Space Environ Med. 2000;71:843-848.
113. Takahashi M, Arito H, Fukuda H. Nurses' workload associated with 16-h night shifts. II:
Effects of a nap taken during the shifts. Psychiatry Clin Neurosci. 1999;53:223-225.
114. Ferrara M, De Gennaro L, Bertini M. The effects of slow-wave sleep (SWS) deprivation
and time of night on behavioral performance upon awakening. Physiol Behav. 1999;68:55-
61.
115. Ferrara M, De Gennaro L, Casagrande M, Bertini M. Selective slow-wave sleep
deprivation and time-of-night effects on cognitive performance upon awakening.
Psychophysiology. 2000;37:440-446.
531

116. Bruck D, Pisani DL. The effects of sleep inertia on decision-making performance. J Sleep
Res. 1999;8:95-103.
117. Ferrara M, De Gennaro L, Bertini M. Time-course of sleep inertia upon awakening from
nighttime sleep with different sleep homeostasis conditions. Aviat Space Environ Med.
2000;71:225-229.
118. Jewett ME, Wyatt JK, Ritz-De Cecco A, Khalsa SB, Dijk DJ, Czeisler CA. Time course of
sleep inertia dissipation in human performance and alertness. J Sleep Res. 1999;8:1-8.
119. Rosekind MR, Smith RM, Miller DL, et al. Alertness management: strategic naps in
operational settings. J Sleep Res. 1995;4:62-66.
120. James M, Tremea MO, Jones JS, Krohmer JR. Can melatonin improve adaptation to night
shift? Am J Emerg Med. 1998; 16:367-370.
121. Jorgensen KM, Witting MD. Does exogenous melatonin improve day sleep or night
alertness in emergency physicians working night shifts? Ann Emerg Med. 1998;31:699-
704.
122. Wrenn K, Wright S. Melatonin after night shift work. Ann Emerg Med. 1999;33:479.
123. Wright SW, Lawrence LM, Wrenn KD, Haynes ML, Welch LW, Schlack HM.
Randomized clinical trial of melatonin after night-shift work: efficacy and
neuropsychologic effects. Ann Emerg Med. 1998;32:334-340.
124. Walsh JK, Muehlbach MJ, Humm TM, Dickins QS, Sugerman JL, Schweitzer PK. Effect
of caffeine on physiological sleep tendency and ability to sustain wakefulness at night.
Psychopharmacology. 1990;101:271-273.
125. Gaba DM. Physician work hours: The "sore thumb" of organizational safety in tertiary
health care. Paper presented at: Proceedings of Enhancing Patient Safety and Reducing
Errors in Health Care, 1998; Rancho Mirage, CA.
532

533

Chapter 47. Safety During Transport of Critically Ill Patients
Susana B. Martins, MD, MSc
Kaveh G. Shojania, MD
University of California, San Francisco School of Medicine
and Institute for Health Policy Studies
Background
The care of acutely ill patients routinely includes transportation, both within a given
hospital to undergo tests and procedures, and between hospitals, as patients may require transfer
to other facilities for specialized services. Critically ill patients in particular commonly require
such transfers and are at high risk for complications en route. 1-4 Developing practices to reduce
or minimize this necessary risk represents a potentially important area of patient safety research.
This chapter focuses on transportation of critically ill patients by health professionals
(paramedics, nurses, physicians and/or respiratory therapists) between hospitals (to receive
higher levels of care) and within the hospital (for diagnostic or therapeutic procedures).
Stabilization before transport, in the field or in the transferring hospital, and the mode of
transferring patients from the field to specialized centers also present important research and
policy questions. 5 However, we regarded these issues as clinical research topics and quality
improvement issues for the fields of pre-hospital and emergency medicine, rather than patient
safety in general, and so do not review this literature here.
Practice Description
Intrahospital transport refers to transportation of patients within a hospital for the
purpose of undergoing diagnostic or therapeutic procedures or transfer to a specialized unit. In
the context of this chapter, this generally involves movement of critically ill patients from
intensive care areas of the hospital (including intensive care units, emergency departments,
operating theaters and recovery rooms) to areas typically not involved in the delivery of such
care (eg, a hospital radiology department). Equipment and staffing used for intrahospital
transport varies by hospital, clinical service and patient acuity. Studies of intrahospital transport
have mainly focused on the adequacy of patient monitoring and ventilator support. The specific
practices evaluated in this chapter include:
• The continued use of mechanical ventilation instead of switching to manual
ventilation. Manual ventilation involves a self-inflating bag with or without a
volumeter, while mechanical ventilation consists of a portable, time-cycled,
volume-constant transport ventilator.
• The use of specialized transfer units during intrahospital transport. The unit is
attached to the patient’s bed and contains all equipment necessary to meet the
patient’s needs (ventilation, monitoring and infusion of drugs) in the ICU and
during transport. The unit works as a stand-alone unit.
Interhospital transport refers to transportation of patients between hospitals by ground or
air ambulance. Interhospital transport teams vary widely in composition, training and experience.
The transport team does not always include a physician; even when a physician is present, his or
her training may not include skills necessary for this task. 6-8 Nurses and respiratory therapists
frequently accompany critically ill patients during interhospital transport. Some paramedics
534

eceive special training in skills necessary for the interhospital transport of critically ill patients. 9
As with physicians, the training of nurses and respiratory therapists assigned responsibility for
interhospital transport varies widely. Equipment used during interhospital transport also varies
widely, 6,7 but the practices evaluated in the literature mainly relate to the use of specialized
transport teams.
Specialized transport teams characteristically receive consistent and high levels of
training and experience in the transportation of critically ill patients, 10-12 compared with teams
assembled ad hoc. Further details of the composition of these teams are presented in connection
with the specific studies reviewed below (see Table 47.1). Because of the relative paucity of
studies of practices for improving the safety of patient transport, we have reviewed the pediatric
and adult literature together.
Prevalence and Severity of the Target Safety Problem
Adverse events during transport of critically ill patients fall into two general categories:
mishaps related to intensive care (eg, lead disconnections, loss of battery power, loss of
intravenous access, accidental extubation, occlusion of the endotracheal tube, or exhaustion of
oxygen supply), and physiologic deteriorations related to critical illness (eg, worsening
hypotension or hypoxemia). Unfortunately many studies do not distinguish clearly between these
2 categories. Further complicating assessments of patient transport as a safety problem is the
confounding effect of patient selection, as patents requiring intra- or interhospital transport likely
represent a sicker patient population than unselected critically ill patients. In fact, one casecontrol
study reported no differences in adverse events (equipment-related or physiologic) in
critically ill adults during the period of intrahospital transportation as compared to matched
subjects in the ICU. 13
Death during transport is a rare event. The majority of studies reported no mortality
during intrahospital transport 13-17 or interhospital transport, 18,19 and some do not mention
deaths.21-23,24
For intrahospital transport of critically ill patients, reported rates of adverse events range
from 5.9% to 66%. 13,14,16,21,22,25 (We could find no comparable reports of event rates for critically
ill children.) Much of this variation undoubtedly reflects definitional differences, but differences
in patient populations also contribute to this wide range. For instance, a prospective study of 50
high-risk adult cardiac patients reported arrhythmias in 84% of patients, with 52% of these
arrhythmias providing an indication for emergency treatment. 17 These event rates are clearly
much higher than would be observed in an unselected population of critically patients. Similarly,
Insel et al showed a significantly higher incidence of hemodynamic changes requiring
therapeutic intervention when intrahospital transport involved transfers from the operating room
to the ICU compared with patients transported from the ICU to diagnostic procedures. 26
In contrast to the above, the literature on adverse events during interhospital transport has
generally involved critically ill children, not adults. Reported rates of adverse events during
pediatric interhospital transport range from 0 to 75%. 2,10-12,19,24,27-29 In one of these studies, a
prospective cohort design reported a morbidity ratio of 1.85 (95% CI: 1.12-3.06) for pediatric
patients transported from another hospital to the pediatric ICU (PICU) as compared with those
admitted directly (emergency room and wards). Importantly, this increased morbidity reflected
an increased rate of “intensive care events” such as plugged endotracheal tubes and loss of
intravenous access, not an increase in physiologic events. Patients experiencing such adverse
events tended to have higher morbidity scores (on the PRISM scale) and lower therapy level
(TISS) scores prior to transport. Thus, as noted above, confounding of differences in patient
535

sickness and intensity of therapy could account for much of the observed variation in transportassociated
morbidity. 24
Opportunity for Impact
A survey conducted in 1990 to review voluntary compliance with the American
Academy of Pediatrics (AAP) recommendations to include physicians with higher level of
training (at least 3 rd year residency) reported that only 28% of hospitals with a pediatric critical
care transport team met this recommendation. All teams included a nurse with pediatric
experience and a varying degree of training, and 50% of teams included a respiratory therapist. 7
Subchapter 47.1. Interhospital Transport
Study Designs and Outcomes
We identified 3 studies with at least a Level 3 study design and Level 2 outcomes (see
Table 47.1). Two of these studies 10,12 involved pediatric patients. One12 reported the prospective
comparison of outcomes for high-risk pediatric patients admitted to two different ICU’s, one of
which employed a specialized transport team, while the other followed the standard practice of
using non-specialized teams. The specialized team consisted of a second-year pediatric resident
and a pediatric ICU nurse, both trained in pediatric advanced life support, and a respiratory
therapist with pediatric experience. Non-specialized teams varied in composition—a physician
was not always present and level of training in pediatric care for other personnel was not
standardized. The other pediatric study, from England, 10 retrospectively compared outcomes
using a specialized team for the transport of intubated newborns from hospitals within 80 miles
to a NICU at a referral center to outcomes during a control period in which transport was
performed by ad hoc doctor/nurse teams. The specialized teams included physicians with more
years of experience and dedicated transport nurses with specialized training, as well as slight
equipment improvements (humidifier for ventilator and invasive/noninvasive blood pressure
monitoring).
The third study (the one involving adults) describes the experience of a London teaching
hospital that receives critically ill patients from other facilities by two methods: either
accompanied by the receiving hospital’s special retrieval team consisting of an ICU physician,
nurse, and medical physics technician (a technician to fix and maintain equipment) or standard
ambulance transport, with an escorting physician supplied by the referring hospital.
The 2 pediatric studies 10,12 reported adverse events during transportation. We counted
adverse events related to intensive care (eg, accidental extubation) as Level 2 and physiologic
events (eg, ph < 7.2) as Level 3. (A case could be made for classifying both types of adverse
events as Level 3, as neither has a clearly established relationship to adverse events of interest).
All studies provided information on case mix in the study and control groups.
Evidence for Effectiveness of the Practice
Although of theoretical and practical concern, the literature to support the scope,
frequency and outcome of adverse events during transportation is sparse and methodologically
weak. Most studies are small descriptive studies of local practices. Factors that limit
comparability between studies include a variety of definitions for transport-related adverse
events, unclear descriptions of transport team training and experience, diverse equipment
536

availability and different scoring systems for severity of illness (APACHE II, APACHE III,
Glasgow Coma Scale, PRISM, etc). Many confounders affect the evaluation of transportation of
a critically ill patient, among them selection bias, the intervention received at primary hospital,
time spent at primary hospital, adequate stabilization before transport and duration of transport.
As shown in Table 47.1, 2 studies involving pediatric populations revealed reductions in
intensive care-related adverse events through the use of specialized teams for interhospital
transport. 10, 12 In one of the studies, patients transported by the standard (non-specialized) team
were older and more likely to have trauma as a diagnosis.12 This difference in patient
populations clearly limits the ability to interpret the results, although the direction of bias this
might introduce is not clear. The other pediatric study 10 reported no significant differences in
basic clinical and demographic factors between the 2 patient populations, but did not report
PRISM scores.
The single study in adults did not report intensive care-related adverse events, but did
observe significant reductions in surrogate physiologic markers and a non-significant reduction
in mortality within 12 hours of arrival at the receiving facility. Although an observational study,
there were no differences in the patient populations in terms of demographic factors, basic
physiology measurements (FiO2, PaO2, PaCO2, PaO2/FiO2, MAP, heart rate and temperature) or
sophisticated measures of severity of illness (APACHE II, Simplified Acute Physiological
Score-SAPS II).
Studies were underpowered to detect significant mortality differences.
Potential for Harm
A delay in the transfer of critically ill patients to referral hospitals because the specialized
team is not available in timely fashion could create a potential for harm although one study
showed no delay or cancellation due to unavailability of specialist team. 11
Costs and Implementation
Although no firm recommendation can be made, the costs and implementation
requirements may only be feasible for tertiary centers that have enough volume to justify the
investment in human and physical resources. Time out of hospital will vary depending on the
time required to stabilize the patient—not the focus of our study. (One study reported an increase
in stabilization time from 80-105 minutes (p

Subchapter 47.2. Intrahospital Transport
Study Designs and Outcomes
As shown in Table 47.2, the 3 studies of manual versus mechanical ventilation employed
a randomized (or quasi-randomized) controlled design. Randomization procedures were not
described in two studies 30, 31 and used the last digit of the patient record in one study. 32 The
quasi-randomized study 32 implemented a crossover design using manual ventilation or transport
ventilator on one leg of the journey and vice-versa on the other leg.
Two studies reported on Level 2 and 3 outcomes 30, 32 and one on level 3 outcomes, 31
venous pressure, oxygen saturation, PetCO2 and mean airway pressure during transport. All
studies report on before-after variation, one study 30 also reported on minute variations during the
first 8 minutes of transport (see Table 47.2). Only one study reported scores for severity of
illness (PRISM). 30 Case-mix was inadequately reported in one study 31 and not reported in
another. 32
It is worth briefly noting a third practice which involves the use of mobile bed/monitor
units versus standard procedure for intrahospital transportation. Studies on this topic are limited
to descriptive experiences of local practice with no definition or systematic evaluation of adverse
events, 17,33-35 so these practices were not reviewed further.
Evidence for Effectiveness of the Practice
The clinical significance of the hyperventilation observed in manually ventilated patients
during intrahospital transportation has yet to be determined. Mechanical ventilation was
associated with respiratory alkalosis when precision of ventilatory settings was inaccurate. No
adverse effect, (ie, morbidity) was observed as a result of the method of ventilation. Use of a
volumeter when manually ventilating a patient reduced the risk of hyperventilation. Studies were
underpowered to detect significant mortality differences.
538

Potential for Harm
Inadequate maintenance and/or precision of transport ventilator may create an opportunity for
harm.
Costs and Implementation
Portable mechanical ventilators are much more expensive, require more hours of training to
manipulate and frequent use to maintain experience. Costs were not mentioned.
Comment
One randomized controlled trial in pediatric postoperative cardiac patients showed an increase in
markers of hyperventilation for patients in the manually-ventilated group. Otherwise, manual
ventilation appears to achieve results comparable to portable mechanical ventilation. Use of a
volumeter when manually ventilating patients, and the addition of a blender to reduce FiO2 when
manually ventilating neonates, 30 may adequately reduce the risks of hyperventilation, rendering
mechanical ventilation during intrahospital transport unnecessary.
539

Table 47.1. Specialized transport teams versus standard care for interhospital transport
Study Setting Study Design,
Outcomes
Critically ill children transported to two
PICU’s in Albany, NY (specialized
transport team) and Syracuse, NY
(standard care): 1992-94 12
Intubated newborns transported by
specialized doctor/nurse team (increased
training and experience) to NICU in
Nottingham, England: 1994-95, and
historical control period during which
non-specialized (ad hoc) doctor/nurse
team transported patients to the same
NICU: 1991-93 10
Critically ill adults transported to a
university ICU in London: 1996-1997;
specialist team with mobile ICU
compared with emergency ambulance
with medical escort 11
Level 3,
Level 2
Level 3,
Levels 2&3†
Level 3,
Levels 1&3
540
Main Results
Significant decrease adverse event
related to intensive care for patients
transported with specialized team
compared to standard care: 1/47 (2%)
vs. 18/92 (20%), p0.05
Nonsignificant reduction in
endotracheal tube-related events
(blocked or dislodged endotracheal
tubes): 0/146 (95% CI: 0-3.2%)
patients transported by specialized
teams vs. 3/73 (4.1%, 95% CI: 1.1-
12.3%) by ad hoc teams
Reductions also observed in adverse
physiologic end-points; such as
abnormal ph (p

Table 47.2 Manual versus mechanical ventilation for intrahospital transportation
Study Setting Study
Design,
Outcomes
30 ventilator dependent,
critically ill adults in ICU:
manually ventilated with
self-inflating bag group,
manually ventilated selfinflating
bag with volumeter,
and mechanically ventilated
group (Federal Republic of
Germany) 31
28 critically ill adults and
adolescents transported from
emergency department for
diagnostic procedures in a
University Hospital: manual
ventilation versus transport
ventilator (US) 32
51 pediatric postoperative
cardiac surgery patients who
were transported within
hospital while intubated:
manually ventilated versus
mechanically ventilated
(US) 30
Level 1,
Level 3
Level 1,
Level 2&3
Level 1,
Level 2&3
Main Results
Pre/post transport: PaCO2 decreased from 41±2 to
34±2 (p

References
1. Pollack MM, Alexander SR, Clarke N, Ruttimann UE, Tesselaar HM, Bachulis AC.
Improved outcomes from tertiary center pediatric intensive care: a statewide comparison of
tertiary and nontertiary care facilities. Crit Care Med. 1991;19:150-159.
2. Kanter RK, Tompkins JM. Adverse events during interhospital transport: physiologic
deterioration associated with pretransport severity of illness. Pediatrics. 1989;84:43-48.
3. Borker S, Rudolph C, Tsuruki T, Williams M. Interhospital referral of high-risk newborns
in a rural regional perinatal program. J Perinatol. 1990;10:156-163.
4. McCloskey KA, Faries G, King WD, Orr RA, Plouff RT. Variables predicting the need for
a pediatric critical care transport team. Pediatr Emerg Care. 1992;8:1-3.
5. Young JS, Bassam D, Cephas GA, Brady WJ, Butler K, Pomphrey M. Interhospital versus
direct scene transfer of major trauma patients in a rural trauma system. Am Surg.
1998;64:88-91; discussion 91-92.
6. Dryden CM, Morton NS. A survey of interhospital transport of the critically ill child in the
United Kingdom. Paediatr Anaesth. 1995;5:157-160.
7. McCloskey KA, Johnston C. Pediatric critical care transport survey: team composition and
training, mobilization time, and mode of transportation. Pediatr Emerg Care. 1990;6:1-3.
8. Gentleman D, Jennett B. Audit of transfer of unconscious head-injured patients to a
neurosurgical unit. Lancet. 1990;335:330-334.
9. Domeier RM, Hill JD, Simpson RD. The development and evaluation of a paramedicstaffed
mobile intensive care unit for inter-facility patient transport. Prehospital Disaster
Med. 1996;11:37-43.
10. Leslie AJ, Stephenson TJ. Audit of neonatal intensive care transport–closing the loop. Acta
Paediatr. 1997;86:1253-1256.
11. Bellingan G, Olivier T, Batson S, Webb A. Comparison of a specialist retrieval team with
current United Kingdom practice for the transport of critically ill patients. Intensive Care
Med. 2000;26:740-744.
12. Edge WE, Kanter RK, Weigle CG, Walsh RF. Reduction of morbidity in interhospital
transport by specialized pediatric staff. Crit Care Med. 1994;22:1186-1191.
13. Hurst JM, Davis K, Johnson DJ, Branson RD, Campbell RS, Branson PS. Cost and
complications during in-hospital transport of critically ill patients: a prospective cohort
study. J Trauma. 1992;33:582-585.
14. Szem JW, Hydo LJ, Fischer E, Kapur S, Klemperer J, Barie PS. High-risk intrahospital
transport of critically ill patients: safety and outcome of the necessary "road trip". Crit
Care Med. 1995;23:1660-1666.
15. Wallen E, Venkataraman ST, Grosso MJ, Kiene K, Orr RA. Intrahospital transport of
critically ill pediatric patients. Crit Care Med. 1995;23:1588-1595.
16. Indeck M, Peterson S, Smith J, Brotman S. Risk, cost, and benefit of transporting ICU
patients for special studies. J Trauma. 1988;28:1020-1025.
17. Taylor JO, Chulay, Landers CF, Hood W, Abelman WH. Monitoring high-risk cardiac
patients during transportation in hospital. Lancet. 1970;2:1205-1208.
18. Macnab AJ. Optimal escort for interhospital transport of pediatric emergencies. J Trauma.
1991;31:205-209.
19. Barry PW, Ralston C. Adverse events occurring during interhospital transfer of the
critically ill. Arch Dis Child. 1994;71:8-11.
542

20. Duke G, Green J. Outcome of critically ill patients undergoing interhospital transfer.
Medical Journal of Australia. 2001;174:122-125.
21. Smith I, Fleming S, Cernaianu A. Mishaps during transport from the intensive care unit.
Crit Care Med. 1990;18:278-281.
22. Andrews PJ, Piper IR, Dearden NM, Miller JD. Secondary insults during intrahospital
transport of head-injured patients. Lancet. 1990;335:327-330.
23. Braman SS, Dunn SM, Amico CA, Millman RP. Complications of intrahospital transport in
critically ill patients. Ann Intern Med. 1987;107:469-473.
24. Kanter RK, Boeing NM, Hannan WP, Kanter DL. Excess morbidity associated with
interhospital transport. Pediatrics. 1992;90:893-898.
25. Kalisch BJ, Kalisch PA, Burns SM, Kocan MJ, Prendergast V. Intrahospital transport of
neuro ICU patients. J Neurosci Nurs. 1995;27:69-77.
26. Insel J, Weissman C, Kemper M, Askanazi J, Hyman AI. Cardiovascular changes during
transport of critically ill and postoperative patients. Crit Care Med. 1986;14:539-542.
27. Bion JF, Wilson IH, Taylor PA. Transporting critically ill patients by ambulance: audit by
sickness scoring. Br Med J (Clin Res Ed). 1988;296:170.
28. Waddell G, Scott PD, Lees NW, Ledingham IM. Effects of ambulance transport in
critically ill patients. Br Med J. 1975;1:386-389.
29. Britto J, Nadel S, Maconochie I, Levin M, Habibi P. Morbidity and severity of illness
during interhospital transfer: impact of a specialised paediatric retrieval team. BMJ.
1995;311:836-839.
30. Dockery WK, Futterman C, Keller SR, Sheridan MJ, Akl BF. A comparison of manual and
mechanical ventilation during pediatric transport. Crit Care Med. 1999;27:802-806.
31. Gervais HW, Eberle B, Konietzke D, Hennes HJ, Dick W. Comparison of blood gases of
ventilated patients during transport. Crit Care Med. 1987;15:761-763.
32. Hurst JM, Davis K, Branson RD, Johannigman JA. Comparison of blood gases during
transport using two methods of ventilatory support. J Trauma. 1989;29:1637-1640.
33. Link J, Krause H, Wagner W, Papadopoulos G. Intrahospital transport of critically ill
patients. Crit Care Med. 1990;18:1427-1429.
34. Holst D, Rudolph P, Wendt M. Mobile workstation for anaesthesia and intensive-care
medicine. Lancet. 2000;355:1431-1432.
35. Hanning CD, Gilmour DG, Hothersal AP, Aitkenhead AR, Venner RM, Ledingham IM.
Movement of the critically ill within hospital. Intensive Care Med. 1978;4:137-143.
543

Section H. Role of the Patient
Chapter 48. Procedures For Obtaining Informed Consent
Chapter 49. Advance Planning For End-of-Life Care
Chapter 50. Other Practices Related to Patient Participation
544

545

Chapter 48. Procedures For Obtaining Informed Consent
Laura T. Pizzi, PharmD
Neil I. Goldfarb
David B. Nash, MD, MBA
Thomas Jefferson University School of Medicine
and Office of Health Policy & Clinical Outcomes
Background
The process of obtaining informed consent, whether a written document or an oral
communication is one means of ensuring that patients understand the risks and benefits of a
treatment or medical intervention. Rooted in medical ethics and codified as a legal principle, it is
based on the assertion that a competent individual has the right to determine what will or will not
be done to him or her. 1 The American Medical Association (AMA) Code of Medical Ethics
establishes informed consent as an ethical obligation of physicians. 2 In addition to being an
ethical obligation of physicians, legislation in all 50 states requires that patients be informed of
all important aspects of a treatment and/or procedures, although the details of these laws and
statutes differ greatly. 2 Failure to obtain adequate informed consent renders a physician liable for
negligence or battery3 and constitutes medical malpractice.
To date, studies of informed consent have not investigated outcomes related to the
adequacy of the communication, insofar as this may impact patient safety. 4,5 Physician-patient
communication styles have been linked to lower rates of malpractice claims. 6 Nonetheless, as
noted by Levinson, malpractice claims do not reflect the actual rate of negligence. While some
have hypothesized that better informed consent could improve the patient-physician relationship,
establish trust, increase patient compliance, and provide information that could reduce medical
error, this has not been shown. 7,8 In the absence of a direct link between adequate informed
consent and the reduction of medical error, the “patient safety outcome” reviewed in this chapter
is the patient’s provision of adequate informed consent.
Practice Description
Informed consent is a process through which a physician informs a patient about the risks
and benefits of a proposed therapy and allows the patient to decide whether the therapy will be
undertaken. 9 It may be received in one sitting, or over a period of time, 7 either orally or in
writing or a combination of the two. Informed consent procedures have been instituted in both
research and clinical medicine. In the former case, Federal regulations establish strict guidelines
for informed consent that are monitored by a special board at each institution (Institutional
Review Board). In addition, risks and adverse events that occur while research is in progress are
followed closely and reported. As such informed consent in the research setting differs greatly
from informed consent in the clinical setting. 10
In clinical practice, formal efforts, such as the signing of a consent form, (presumably
preceded by adequate exchange of information), are only undertaken in some circumstances,
notably prior to major invasive procedures such as radiologic procedures and surgery. Less well
appreciated is that all medical care, including pharmacy prescriptions or laboratory tests, requires
informal informed consent, except when the patient is incompetent to make a decision or
relinquishes the right to provide it. 3 Studies suggest that in practice only minimal formal efforts
are made to obtain informed consent for routine interventions. 11, 12
546

Legislation governing the requirements of, and conditions under which, consent must be
obtained varies greatly from State to State. General guidelines, such as those proposed by the
AMA require patients to be informed of the nature of their condition and the proposed
procedure, the purpose of the procedure, the risks and benefits of the proposed treatments, the
probability of the anticipated risks and benefits, alternatives to the treatment and the associated
risks and benefits, and the risks and benefits of not receiving the treatment or procedure. 2,3,13
As discussed below, procedures to obtain informed consent may not adequately promote
the patient’s comprehension of the information provided, rendering the consent not truly
“informed.” Interventions that may prove beneficial in improving and ensuring the patient‘s
understanding include redrafting of consent forms to reduce complexity, providing written
materials to accompany oral conversations, using multimedia or other techniques to improve
comprehension, and asking patients to recap discussions about the procedure.
Prevalence and Severity of Target Safety Problem
Procedures to obtain consent must ensure that the patient understands his or her
condition, as well as the risk and benefits of treatment, and its alternatives. It has been estimated
that less than half of the US population understands commonly used medical terms. 14,15 This
“health literacy” problem may impact the ability of the patient to understand any attempts to
obtain information. In addition to lack of comprehension, procedures to obtain informed consent
may be incomplete.
Several studies have noted the various insufficiencies in procedures to obtain informed
consent. Three studies examined the completeness of physician-patient conversations in
obtaining informed consent. Braddock et al 12 focused on outpatient discussions. Recognizing
that some procedures may require more discussion than others, they created a three-tiered
evaluation procedure, in which the completeness of patient-physician discussion differ according
to the complexity of the decision being discussed. Basic decisions, such as laboratory tests,
require the least in-depth discussion, covering only the patient’s role, the clinical nature of the
decision, and exploration of patient preferences. Intermediate decisions, such as changes in
medication, require a moderate depth of discussion, incorporating the Tier 1 subjects, and adding
a discussion of alternative treatments, the risks and benefits of the alternatives, and an
assessment of the patients understanding. Complex decisions such as surgery require a
discussion of the uncertainties associated with the decision, in addition to all of the
aforementioned steps. Analyzing audiotaped conversations between 1057 patients and 59
primary-care physicians and 65 surgeons, Braddock et al found that 17.2% of basic decisions
contained all required components, while none of the intermediate decisions and only one of the
complex decisions contained all of the required components. Applying only the Tier 1 basic
consent standards, 20.5% of basic decisions, 21.9% of intermediate decisions, and 38.2% of
complex decisions met all the criteria.
In a study of informed consents of surrogates for pediatric patients undergoing surgery,
coders examined audiotaped conversations with surrogates, as well as structured interview and
questionnaire data regarding the conversations. They noted that patients’ recall of the
conversations with physicians often omitted key components, such as the risks and benefits of
the procedures. 16
Bottrell et al 13 examined the completeness of 540 consent forms from 157 hospitals
nationwide. Of these, 26.4% included all four of the basic elements (risks, benefits, alternatives,
and other important aspects of the procedure). Eighty-seven percent noted the general possibility
of risk, but less than half provided specific information. Alternatives were noted in 56.9% of the
547

forms, and benefits appeared in 37%, though most of these were general references rather than
specific information. Although 74% of consent forms were deemed incomplete, it is unknown
whether physician-patient discussions that preceded the signing of the consent form included the
missing information.
A study by Mark et al9 found that 82.4% of 102 participants reported that they
understood everything that their physicians had described about a procedure and indicated that
all of their questions had been answered. Eighteen patients had remaining unanswered questions.
Half of this group requested more time to speak with their physicians, while the other 9 felt that
their questions were not important.
In a study by Lavelle-Jones, 17 69% of patients admitted that they did not read a consent
form before signing it. In addition, approximately half of the patients awaiting treatment were
unhappy with the amount of information they received, with 21% stating that most of the
information they obtained about their surgical treatment was obtained outside of the hospital.
Consent forms have been targeted for their lack of readability. Patients with limited
reading ability are at increased risk for medical errors, due to problems reading medication
bottles, appointment slips, self-care instructions, and health education brochures. 18 These
patients may also have trouble reading materials intended to aid in obtaining informed consent.
According to the National Adult Literacy Survey of 1993, approximately 40-44 million
Americans were functionally illiterate, defined as the inability to complete basic reading tasks
required to function as a member of society. 19 In addition, even in educated adults, the highest
grade-level completed may not reflect actual reading comprehension level. A study of 100 adult
cancer patients found that most read at a mean grade-level equivalent of between 10 th and 11 th
grade. The authors suggest that forms and educational materials be written at a grade-level three
levels below the highest level of education completed by a patient. 20
Several studies have examined the readability of procedure consent forms. Two studies
examined consent forms for radiologic procedures using computer generated “readability”
scores. Consent forms for use with iodinated contrast media found that 12.35 years of education
were required to read consent forms. 21 A similar study of general radiologic procedure consent
forms found that they required a mean of 15 years of education. Only 16% of forms could be
understood by patients with a high-school education. 11 Another Hopper et al study22 found that
general hospital consent forms were written at a grade level of 12.6. Just over half could be
understood by a patient with a high school education, less than a third by patients with a 10 th
grade reading level, and just over 5% by patients with an 8 th grade reading level.
Opportunities for Impact
While informed consent is a well-established practice, it often fails to meet its stated
purpose. Several methods of improving the procedures of obtaining informed consent have been
proposed, including improving the readability of consent forms, 22 asking patients for recall to
establish understanding, 3,23 adding additional stimuli, such as multimedia presentations 24 and
providing written information. 17
Lavelle-Jones 17 found that elderly patients (over 60 years of age) had poorer recall than
younger patients. In addition, patients with internal locus of control—those who believed their
health was in their own control—were better informed than those with an external locus of
control. Patients with above average IQ exhibited better recall. These findings could indicate “atrisk”
groups that interventions may target.
One author argues that an important opportunity for impact is to change the model of
implementing informed consent from a single event approach to a process approach. 7 Currently
548

obtaining informed consent often revolves around the signing of the consent form. Although this
approach clearly delineates the responsibility of health care providers, provides documentation
of the consent, and fits easily into the current provider structure, it often results in patients failing
to actually comprehend the information and reinforces physicians’ conception that the consent
ritual is futile. In contrast, a process model involves the physician providing information over
time, establishing a better patient-physician relationship and better comprehension of medical
care. As of yet there is no data to support these suppositions.
Since the definition of adequate informed consent is debatable, the number of individuals
currently not receiving interventions to obtain adequate informed consent is likely to be quite
high, but is not known.
Study Design, Outcomes and Effectiveness of Various Approaches
Improving Readability of Consent Forms and Education Materials
In order to ensure that patients understand the procedure to which they are consenting, it
is important that all materials be presented in a comprehensible manner. Consent forms are
written with relatively complex sentence structure and vocabulary, making it difficult for the
average adult to interpret the information. In addition, providing consent forms in the primary
language of patients may improve consent procedures. However, we located no studies
examining the effectiveness of such practices.
Structured Discussions
Informed consent is often obtained during informal discussion between physicians or
nurses and patients, and (as discussed above), these discussions frequently do not cover all of the
relevant information. 12 Two studies examined the use of a structured interview format in
providing information to patients. Solomon et al 25 studied 36 patients receiving cardiac
catheterization for the first time at a Veterans hospital. Patients were randomized into two groups
(Level 1 design). Both groups were briefed by a cardiologist regarding the procedure (standard
care). In addition, the experimental group received a 30-minute structured teaching session with
a nurse to discuss all aspects of the procedure, including the purposes, techniques, sensations,
risks and benefits. Patients in the experimental group also received an illustrated guide to
cardiovascular procedures and an educational pamphlet. All subjects were tested using a 13-item
questionnaire covering the information that should have been imparted during informed consent
procedures (Level 3 outcome). The intervention group scored significantly better than the control
group (11.5 vs. 8.9).
In a later study, Dawes et al 26 studied 190 hospitalized patients undergoing ENT surgery.
Patients were assigned to one of 4 groups. Groups 1 and 2 were assigned at different times
(Level 2 design), while groups 3 and 4 were randomly assigned (Level 1 design). Group 1 had
no consent interview until after assessment at the conclusion of the study. Group 2 engaged in an
informal interview with a physician (reflecting current practice), during which the number of
complications discussed was recorded. Group 3 engaged in a structured interview with a
physician, covering the purpose and technique of the procedure, complications, sensations,
alternatives, and benefits (a checklist was used as a guide). The final group, Group 4, engaged in
the same structured interview, except that the patient was provided with a copy of the checklist
and allowed to take it with them after the interview. Patient anxiety was assessed following
consent using a visual analog scale, then reassessed at an interview, given 4 hours later. At the
later interview, patients were asked to recount orally the operation name, describe what would be
549

done, list complications, and state whether they understood the information given to them (Level
3 outcome). All but Group 1 (control group) patients showed a drop to normal anxiety after
informed consent. Group 2 (informal discussion) remembered proportionately more
complications mentioned in the informal interview, although fewer complication were covered.
Groups 3 and 4 recalled more total complications than Group 2. There was no differences
between Groups 3 and 4 (structured interviews).
Asking for Recall
A simple method of determining whether a patient understands information regarding a
procedure is to ask the patient to recount what he or she has been told. 3 Two studies examined
this intervention, using a randomized controlled trial design (Level 1 design). In the first study,
informed consent was solicited from 50 patients undergoing percutaneous lung biopsy. 23
Twenty-seven patients in the control group were offered the standard procedure for obtaining
informed consent: approximately 30 minutes prior to the procedure, the physician described the
procedure in detail, including its risks and benefits. Four complications were specifically
described, along with their relative risks. Patients were asked to sign a standard consent form.
Twenty-three patients received the same procedure, but in addition, they were asked to describe
all 4 potential complications. The procedure was repeated until all patients in the intervention
group could recount all of the complications. This modified procedure usually took less than 5
minutes extra compared with the traditional approach. Patients were interviewed 2 hours after
the procedure was completed. Patients in the modified consent group had better recall (Level 2
outcome) than patients in the control group (56% vs. 14% with high recall (recalling 3-4 out of 4
risks), and 13% vs. 44% with low recall (recalling 0-1 out of 4 risks)).
A second study examined verbalization in 20 patients undergoing anterior cruciate
ligament (ACL) reconstruction. 27 Patients were randomly assigned to 2 groups. Both groups
received the standard education for ACL reconstruction, which included the use of a 3-D model
of the knee, discussion with a physician about the procedure, and obtaining informed consent.
The experimental group (8 subjects) also were asked to repeat back the risks of the procedure
until they could accurately recall all risks discussed. One-month later, recall regarding the
information received was tested using a 3-item questionnaire (Level 2 outcome). All 8 in the
experimental group answered all questions correctly, while only four out of the 12 in the control
group answered all questions correctly (p=0.03).
Use of Visual or Auditory Learning Aids
Adding additional stimuli may increase the ability of the patient to understand
information being conveyed or increase retention of that information. For instance, use of visual
diagrams may make a procedure easier to understand. Multi-media may also promote better
comprehension. Three studies have examined the addition of visual stimuli for informed consent.
One study of patients undergoing back surgery examined the impact of a diagnosis-specific
videodisk program on patient outcomes. 24 Patients (n=393) who were candidates for elective
back surgery (primarily for herniated disc and spinal stenosis) were randomized to two education
groups (Level 1 design). The control group received a written booklet regarding the surgical and
non-surgical treatments for herniated disc and spinal stenosis, a description of expected
outcomes, and a short self-test on the booklet information. Patients in the experimental group
also received the booklet, but in addition viewed a videodisk program. The program allowed a
patient to enter their diagnosis and age and receive customized information on the alternative
treatments, discussion of the diagnoses (and the ambiguities of diagnosis), and interviews from
550

patients that had been treated surgically and non-surgically. At the end the patient was provided
with a printout of outcome probabilities. Patients in the experimental group were slightly more
likely to report that they had all the information they wanted (Level 3 outcome). Rates of
patients consenting to surgery for herniated disk (32% vs. 47%) and other diagnoses (5.4% vs.
14.0%) were lower in the videodisk group.
Hopper et al28 also tested an interactive video program, although the tested program was
computer based. One-hundred and sixty outpatients referred for IV contrast media studies were
stratified by age, sex, and previous exposure to contrast media, then randomized to receive either
a written consent form, or an interactive computer-based video (Level 1 design). Subjects in the
control group received a consent form designed to be read at an eighth grade reading level.
Subjects in the experimental group viewed a video in which a physician used identical words as
the consent form to inform subjects about the procedure and risks. Subjects then had an option of
hearing more about the risks. If subjects chose not to hear additional information, they were
provided with printouts of the risks (with the minimal information already given). Otherwise
subjects were provided with printouts certifying that they had completed the program. All
subjects were tested using a 7-item questionnaire regarding the procedure and risks (Level 2
outcome). Patients that viewed the video responded correctly more often than the control group
when asked about general aspects of the procedure. Female patients in the video group also
responded correctly more often to questions about risks, although this finding did not hold for
male patients. The video did take approximately 1.6 minutes of additional time to complete, and
there was no difference in patients desire for additional knowledge between the groups (Level 3
outcome).
One final study did not use an interactive video, but tested whether providing information
via video was superior to providing information via an informal discussion with the physician
(standard practice). 29 Two-hundred and twenty four subjects referred for colonoscopy were
stratified into previous or no previous colonoscopy, then randomized into 3 groups (Level 1
design). The control group had a structured discussion with a physician, in which the physician
covered the same information covered in the video according to a checklist. The video-only
group watched a 5-minute videotape, in which a physician described the procedure, and its risk
and benefits. The third group (video and discussion), watched the video and then engaged in a
structured discussion with a physician. All patients were tested using a 13-item questionnaire
(Level 2 outcome). Both video groups gave more correct responses than the discussion-only
group, although they did not differ from each other. In addition, patient anxiety levels were
measured using the State-Trait Anxiety Inventory (Level 3 outcome). There were no differences
among the 3 groups.
Providing Written information to Patients
Providing written information to patients regarding their diagnoses, proposed treatments,
and other information given during informed consent discussion allows the patient to refer back
to such information, and possibly increases comprehension. One early study compared the thencommon
practice, informal interview between patient and doctor, with provision of a written
consent form. 30 Eighty patients, referred for a first excretory urography, were randomly assigned
to 2 groups (Level 1 design). The control group received standard care, the informal interview.
The experimental group received the same interview along with a detailed written consent form
to read and sign. Patient clinical reactions and side effects were monitored (Level 1 outcome).
One to 3 days after the procedures, patients retrospectively rated discomfort, fear or
apprehension, and understanding of the procedure (Level 3 outcome). Patients also completed an
551

8-item knowledge examination covering the information on the consent form (Level 2 outcome),
or in the informal interview. Experimental subjects scored significantly better (p=

References
1. Edwards SJ, Lilford RJ, Thornton J, et al. Informed consent for clinical trials: in search of
the "best" method. Soc Sci Med. 1998;47:1825-40.
2. American Medical Association. Informed consent. American Medical Association. April
11, 2001.
3. Pape T. Legal and ethical considerations of informed consent. AORN J. 1997;65:1122-7.
4. Stewart MA. Effective physician-patient communication and health outcomes: a review.
CMAJ. 1995;152:1423-33.
5. Greenfield S, Kaplan S, Ware JE, Jr. Expanding patient involvement in care. Effects on
patient outcomes. Ann Intern Med. 1985;102:520-8.
6. Levinson W, Roter DL, Mullooly JP, et al. Physician-patient communication. The
relationship with malpractice claims among primary care physicians and surgeons.
7. Lidz CW, Appelbaum PS, Meisel A. Two models of implementing informed consent. Arch
Intern Med. 1988;148:1385-9.
8. Kuyper AR. Patient counseling detects prescription errors. Hosp Pharm. 1993;28:1180-1,
1184-9.
9. Mark JS, Spiro H. Informed consent for colonoscopy. A prospective study. Arch Intern
Med. 1990;150:777-80.
10. Code of Federal Regulations. Title 45: Public Welfare. .Revised June 18, 1991.Effective
August 19, 1991.
11. Hopper KD, TenHave TR, Hartzel J. Informed consent forms for clinical and research
imaging procedures: how much do patients understand? AJR Am J Roentgenol.
1995;164:493-6.
12. Braddock CH, 3rd, Edwards KA, Hasenberg NM, et al. Informed decision making in
outpatient practice: time to get back to basics. JAMA. 1999;282:2313-20.
13. Bottrell MM, Alpert H, Fischbach RL, et al. Hospital informed consent for procedure
forms: facilitating quality patient-physician interaction. Arch Surg. 2000;135:26-33.
14. Boyle CM. Difference between patients' and doctors' interpretation of some common
medical terms. Br Med J. 1970;1:286-9.
15. Cole R. The understanding of medical terminology used in printed health education
materials. Health Educ J. 1979;38:111-21.
16. Lashley M, Talley W, Lands LC, et al. Informed proxy consent: communication between
pediatric surgeons and surrogates about surgery. Pediatrics. 2000;105:591-7.
17. Lavelle-Jones C, Byrne DJ, Rice P, et al. Factors affecting quality of informed consent.
BMJ. 1993;306:885-90.
18. Baker DW, Parker RM, Williams MV, et al. Health literacy and the risk of hospital
admission. J Gen Intern Med. 1998;13:791-8.
19. Kirsch I, Jungblut A, Jenkins L, et al. Adult literacy in America: a first look at the results
of the national adult literacy survey. Washington, DC: National Center for Education, US
Department of Education; 1993.
20. Jubelirer SJ, Linton JC, Magnetti SM. Reading versus comprehension: implications for
patient education and consent in an outpatient oncology clinic. J Cancer Educ. 1994;9:26-
9.
21. Hopper KD, Lambe HA, Shirk SJ. Readability of informed consent forms for use with
iodinated contrast media. Radiology. 1993;187:279-83.
553

22. Hopper KD, TenHave TR, Tully DA, et al. The readability of currently used
surgical/procedure consent forms in the United States. Surgery. 1998;123:496-503.
23. White CS, Mason AC, Feehan M, et al. Informed consent for percutaneous lung biopsy:
comparison of two consent protocols based on patient recall after the procedure. AJR Am J
Roentgenol. 1995;165:1139-42.
24. Deyo RA, Cherkin DC, Weinstein J, et al. Involving patients in clinical decisions: impact
of an interactive video program on use of back surgery. Med Care. 2000;38:959-69.
25. Solomon J, Schwegman-Melton K. Structured teaching and patient understanding of
informed consent. Crit Care Nurse. 1987;7:74-9.
26. Dawes PJ, O'Keefe L, Adcock S. Informed consent: the assessment of two structured
interview approaches compared to the current approach. J Laryngol Otol. 1992; 106:420-4.
27. Wadey V, Frank C. The effectiveness of patient verbalization on informed consent. Can J
Surg. 1997;40:124-8.
28. Hopper KD, Zajdel M, Hulse SF, et al. Interactive method of informing patients of the risks
of intravenous contrast media. Radiology. 1994;192:67-71.
29. Agre P, Kurtz RC, Krauss BJ. A randomized trial using videotape to present consent
information for colonoscopy. Gastrointest Endosc. 1994;40:271-6.
30. Winfield AC, Ford CV, James AE, et al. Response of patients to informed consent for
excretory urography. Urol Radiol. 1986;8:35-9.
31. Neptune SM, Hopper KD, Houts PS, et al. Take-home informed consent for intravenous
contrast media: do patients learn more? Invest Radiol. 1996;31:109-13.
554

555

Chapter 49. Advance Planning For End-of-Life Care
Nina Garas, MD
Emory University School of Medicine
Steven Z. Pantilat, MD
University of California, San Francisco School of Medicine
Background
Physicians and other health care workers have long struggled with decisions regarding
care for patients at the end of life. An important component of this care involves assessing and
understanding patient preferences for care through ongoing discussions with competent adult
patients and/or their family members or surrogates. Advance care planning protects patient
autonomy and helps to assure that their health and medical treatment wishes are implemented.
Good communication at the end of life can also help patients achieve closure and meaning in the
final days of their life.
Over the past 20 years, public consciousness regarding planning for end-of-life care has
been raised through several seminal court cases, such as those involving Karen Ann Quinlan and
Nancy Cruzan. These cases and the public interest they helped engender led to legislation
promoting patients’ rights to determine their care at the end of life. For example, Natural Death
Acts (statutes passed by State legislatures that assert a person’s right to make decisions regarding
terminal care) have helped promote the use of living wills (described below). 1 In addition, in
1990 the Federal Patient Self-Determination Act (PSDA) was passed by Congress to encourage
competent adults to complete advance directives. The PSDA requires hospitals, nursing homes,
health maintenance organizations, and hospices that participate in Medicare and Medicaid to ask
if patients have advance directives, to provide information about advance directives, and to
incorporate advance directives into the medical record. 2
Advance directives are any expression by a patient intended to guide care, should they
lose their medical decision making capacity. Although both oral and written statements are valid,
the added effort required to complete written statements gives them greater weight. In addition to
their use when patients lose competence, advance directives also help patients consider the type
of care they would want in the future, even if they retain decision making capacity. Advance
directives have legal validity in almost every State.
There are 2 principal forms of written advance directives: living wills and durable powers
of attorney for health care. A living will is a document that allows an individual to indicate the
interventions he or she would want if he or she is terminally ill, comatose with no reasonable
hope of regaining consciousness, or in a persistent vegetative state with no reasonable hope of
regaining significant cognitive function. A durable power of attorney for health care (DPOA-
HC) is a more comprehensive document that allows an individual to appoint a person to make
health care decisions for him or her should he or she lose decision making capacity.
Prevalence and Severity of the Target Safety Problem
Respecting patient preferences regarding end-of-life care requires a well-coordinated
approach. Problems can arise in both documenting patient preferences and ensuring that
preferences are available and respected at the time they are needed. In addition, inadequate
communication with patients can compromise the goal of respecting patient preferences for endof-life
care through a variety of mechanisms.
556

Failure to Document Preferences
The PSDA was a legislative solution (see Chapter 55) designed to increase rates of
completed advance directives. Although there was initial hope that PSDA would markedly
increase rates of advance directive documentation, by the early 1990s it was clear that the impact
was small. At that time, a large multicenter randomized trial, the Study to Understand Prognoses
and Preferences for Outcomes and Risks of Treatments (SUPPORT), was undertaken to improve
advance care planning. SUPPORT represents one of the largest and most comprehensive efforts
to describe patient preferences in seriously ill patients, and to evaluate how effectively patient
preferences are communicated. SUPPORT cost 28 million dollars and enrolled 9100 seriously ill
patients. In SUPPORT, a trained nurse facilitator provided prognostic information to patients and
medical staff, discussed patient preferences with patients and families, and facilitated
communication between patients and physicians.
Neither the PSDA legislation nor the SUPPORT intervention had major impacts on the
documentation of patients’ preferences regarding end-of-life care. Teno et al reported on the
documentation of advance directives at 3 points: before PSDA, after PSDA, and after the
SUPPORT intervention. The percentage of patients with an advance directive was unchanged in
all 3 groups, but documentation of those directives increased at each stage, from 6% to 35% to
78% in the SUPPORT intervention group. Despite this increase in documentation, only 12% of
patients with an advance directive had talked with a physician when completing the document
and only 25% of physicians were aware of their patients’ advance directives. 3 SUPPORT found
that only 23% of seriously ill patients had talked to their doctors about their wishes concerning
cardiopulmonary resuscitation (CPR) and that patient-physician discussions and decisions were
uncommon even in seriously ill patients whose death was predictable. 4 Another study that
surveyed elders in community settings found that the vast majority (81%) stated their desire to
discuss their preferences with their physicians if they were terminally ill, but only 11% had done
so. 5 As these studies demonstrate, patients often want to talk about death and dying but expect
physicians to bring up the issues. 6
Ensuring that Preferences are Available and Respected
Even when advance directives are prepared, studies show they often do not change
interventions at the end of life. 3,7 Advance directives are frequently not available, recognized or
applied, nor do they help reduce hospital resource use. There are multiple reasons why advance
directives may go unrecognized. 8 Admitting clerks may fail to document or incorrectly
document the status of a directive on admission to the hospital. Patients and families often do not
inform the hospital physician or admitting clerk about their advance directives, or fail to bring
documentation to the hospital. 8 In one survey of 200 patients, only 18% had filled out an
advance directive and of these, 50% had secured the only copy in a safety deposit box! 9 A copy
of the advance directive is often not transferred from the nursing home to the hospital on
admission. In a study by Morrison, physicians documented advance directives or discussions
with appointed proxies about treatment decisions in only 11% of admission notes. 8
Although the goal of advance directives is to ensure that patients receive treatment that is
consistent with their preferences, to date there is no evidence that documenting advance
directives leads to this outcome. In SUPPORT, there was no evidence that increasing the rates of
advance directives resulted in care more consistent with patients' preferences. 10 This finding was
concordant with a study of nursing home patients and their family members regarding
preferences for aggressive treatment at the end of life. There, 25% of patients received care that
557

was inconsistent with their previously expressed wishes. 11 The problem may not be the substance
of advance directives per se, but rather in the manner in which clinicians approach them.
Physicians may be hesitant to initiate discussions of advance directives with patients, especially
early in the course of an illness. 12
Despite these shortcomings, advance directives remain the best available approach for
helping patients plan future care. These discussions, difficult as they are, help ensure that
patients receive care consistent with their values and goals, spare the patient inappropriate
interventions, and help maintain dignity during the dying process.
Physician Communication
In order to improve the quality of end-of-life care, physicians need to effectively
communicate with their patients and understand their preferences for care. Several studies have
documented imperfections in physician-patient communication. 13,14 Several studies have
demonstrated that physicians often misunderstand or are unaware of their patients’ preferences
for care. 15,16 Furthermore, physician prediction of patients’ preferences for resuscitation are no
better than random. 3,14
In summary, the published literature demonstrates significant problems in all areas
crucial to advance care planning and ascertainment of patient preferences, transmission of
information to appropriate care settings, and respecting those preferences. The provision of
unwanted end-of-life care is an adverse event that can potentially be avoided by the
implementation of effective patient safety practices.
Opportunities for Impact
Patients with chronic or life-limiting illnesses make up a large proportion of the adult
primary care population. Almost three-quarters of the 2.3 million Americans that die each year
are 65 years of age or older. By the year 2030, people older than 65 will compromise 20% of the
total population (70 million people), compared with 13% in 1994. Today’s average life
expectancy is 75.5 years, and the leading causes of death are heart disease, cancer and stroke.
Data from 1995 estimated that these causes accounted for 62% of all deaths and 67% of deaths
for those age 65 and over. 17 The overall picture is of an aging population, with many individuals
living for several decades (often with chronic diseases) after the possibility of death becomes
more than theoretical. 18
SUPPORT documented serious problems with terminal care. Physicians did not
implement patients’ refusals of interventions. When patients wished to forgo CPR, a do not
resuscitate order was never written in about 50% of cases. 4 While 90% of Americans say they
want to die at home, 4 out of 5 die in a hospital or other health care facility. The SUPPORT
study showed that only 35% of the study patients had an advance directive. These patients had
an approximate six month mortality rate of 50%.
Physicians and the public also commonly overestimate the effectiveness of CPR. In
reality, in-hospital cardiac arrests have a survival rate of about 15%. For patients over 65 the
survival rate is about 10-11%, and 3.5% for patients over age 85. 19 Elderly nursing home patients
with out-of-hospital arrest only have 1-2% survival. 20 Studies have shown that when patients are
aware of the real survival rates for CPR, they are less likely to desire this intervention.
558

Evidence for Effectiveness of the Practice
Documenting Preferences and Ensuring that they are Available and Respected
A Physician Order form for Life-Sustaining Treatment (the POLST)
In the mid-1990s, a task force of ethicists and clinicians at the Oregon Health Sciences
University developed a new Do Not Resuscitate (DNR) order form called POLST (Physician
Orders for Life-Sustaining Treatment). POLST is a comprehensive two-page order form that
documents a patient’s preference for life-sustaining treatments. The form is designed to record a
patient’s wishes clearly and simply. The accompanying POLST wallet card is included as Figure
49.1; the complete POLST form and materials can be obtained from the Oregon Health Sciences
University’s Center for Ethics in Health Care (http://www.ohsu.edu/ethics/polst.htm).
Tolle et al examined the extent to which POLST ensured that nursing home residents’
wishes were honored for DNR orders, and for hospital admission only if comfort measures
failed. None of the 180 patients who completed POLST received CPR, ICU care, or ventilator
support, and only 2% were hospitalized to extend life. The study subjects had low rates of
transfer for aggressive life-extending treatments and high levels of comfort care. 21
Since 1995, more than 220,000 copies of POLST have been distributed throughout the
State. Data from 1999 suggest, albeit circumstantially, that this initiative may be working. In
1996, Oregon’s in-hospital mortality rate was 31%, compared with the national average of
56%. 22
Lee et al studied the effectiveness of POLST in a Program of All-Inclusive Care for the
Elderly (PACE) in Portland, Oregon. They retrospectively reviewed POLST instructions for each
of the 58 participants and whether or not each of the treatments addressed by the POLST was
administered in the final 2 weeks of life. The POLST specified DNR for 50 participants (93%);
CPR use was consistent with these instructions for 49 participants (91%). The participants also
designated the level of care they preferred as either comfort care, limited, advanced, or full
intervention. Interventions administered were at the level specified in only 25 cases (46%), with
less frequent deviations in antibiotic administration, administration of IV fluids, and placement
of feeding tubes. The investigators concluded that the POLST effectively limits the use of some
life-sustaining interventions, but that further investigation is needed into the factors that lead
physicians to deviate from patients' stated preferences about other treatments. 23
559

Administrative Initiatives to Ascertain Preferences on Admission to Hospital or Nursing
Home
In addition to POLST, some medical centers have developed admission order forms to
document patient preferences regarding end of life. These forms require health care personnel to
inquire about advance directives, resuscitation preferences, artificial fluids and nutrition, etc.
This approach, promoted by the passage of the PSDA, may be effective in promoting providerpatient
discussions about end-of-life wishes and prevent unwanted treatments. However, there
are no data documenting the effectiveness of this strategy.
Ascertaining Preferences in the Outpatient Setting
As with other forms of computerized decision support (Chapter 53), computer-generated
reminders for primary caregivers can increase the rates of discussion of advance directives and
completion of advance directive forms among elderly outpatients with serious illnesses. Dexter
et al performed a randomized, controlled trial to test the effectiveness of computerized
reminders. The participants were 1009 patients and 147 primary care physicians in an outpatient
setting. Physicians that received computer-generated reminders that recommended discussion of
one or both of 2 types of advance directives were compared with physicians who received no
reminders. Physicians who did not receive reminders (controls) discussed and completed
advance directives in only 4% of the patients On the other hand, physicians who received both
types of reminders discussed (24%) and completed (15%) advance directives significantly more
frequently. 24
The Portability of Advance Directives between Hospitals and Nursing Homes
Ghusn et al retrospectively studied the relationship between inter-institutional
communication and continuity of advance directives from hospital to nursing home settings.
Having a hospital discussion about advance directives or having a hospital DNR order were
associated with a higher rate of advance directive discussions in nursing homes. Hospital DNR
orders were continued for 93% of patients discharged to the hospital-affiliated nursing home and
41% of patients discharged to the community nursing home. Specific communication of hospital
DNR status to the receiving nursing homes was associated with better continuity of DNR orders.
The authors concluded that completing advance directives before patients are discharged to
nursing homes, communicating advance directives to the receiving home, and providing followup
discussions at the nursing home might improve the continuity of advance directives between
hospitals and nursing homes. 25
Practices to improve physician-patient communication and physician understanding of patient
preferences
Training for Physicians
Physician education is an attractive way to improve end-of-life care. Physicians often do
not communicate about advance care planning because many have not been taught the relevant
communication skills and have learned them only through personal experience. 26 A study by
Tulsky et al revealed that when physicians discussed end-of-life issues with their patients, they
spoke twice as much as they listened and did not routinely explore patients’ values. 26
Until recently, training for health care providers in palliative care and respecting patient
preferences, and materials to support such training, were inadequate. For example, recent studies
560

have demonstrated that most medical and nursing textbooks insufficiently cover end-of-life care
issues. 27 Increasingly, resources (including textbooks, palliative care journals or journal series, 28
Web sites and training programs) are filling this educational void. The American Medical
Association has developed an extensive physician training program titled Education for
Physicians on End-of-Life Care (EPEC). 29 This curriculum teaches fundamental skills in
communication, ethical decision making, palliative care, pain and symptom management, and
other end-of-life treatment issues. 30 The Robert Wood Johnson Foundation initiative, “Last
Acts,” is another ambitious effort to educate both patients and providers.
Other educational training programs exist for physicians and students as well. 30
Physicians can receive formal training by attending conferences on decisions near the end of life,
case management meetings regarding individual patients, and seminars on communication skills
with individual feedback to physicians on their performance. 31 Physicians with expertise in this
area often conduct seminars to educate physicians. Buckman and Lo have developed guides for
specific end-of-life discussions, such as breaking bad news and the act of active listening and
empathy.
32, 33
As attractive as these educational programs are, none have been studied for their impact
on changing practice or outcomes. Although common sense might tell us that such programs are
likely to be effective, the generally unimpressive relationship between professional education
and outcomes or process change (Chapter 54) provides grist for uncertainty pending formal
effectiveness studies.
Palliative Care Services
Specialized palliative care programs have become increasingly common in the health
care system. Physicians and other health care providers, including nurses, social workers,
chaplains, and others are available to coordinate care and provide consultation for terminally ill
patients in hospices, hospitals, nursing homes or patient’s homes. The palliative care service also
plays an important role in fostering communication among providers, patients, and families.
Data regarding effectiveness are lacking.
Hospitalist Systems
Hospitalist physicians may improve end-of-life care in hospitals. Hospitalists, by virtue of their
large inpatient volumes, should become increasingly facile with ascertaining patient preferences
regarding end-of-life care. Hospitalists have a unique opportunity to approach patients, since an
admission generally signals either a worsening of the patient’s current condition or a new
diagnosis. The hospitalist may have more time to spend with patients and is available over
consecutive hospital days to answer any questions. A routine discussion of advance directives by
hospitalists can help improve the quality and efficiency of patient care. 34 On the other hand,
patients may have a long-standing trusting relationship with their primary care physicians, and
may have expressed their wishes to this physician prior to hospitalization. This possibility
highlights the importance of hospitalist-primary care provider communication, particularly
concerning end-of-life issues. 34
One retrospective chart review study of 148 patients dying at a community teaching
hospital has examined the impact of hospitalists on end-of-life care. In this study, patients cared
for by hospitalists were significantly more likely to have had a documented family meeting (91%
vs. 63% for patients of community-based primary physicians). About two-thirds of patients in
both groups requested limitations in the level of care by the time of death. Of these, patients of
hospitalists were significantly less likely to have documented pain, dyspnea, or anxiety in the 48
561

hours prior to death (57% vs. 75%). Whether these differences reflect differences in the quality
of care, the completeness of documentation, or underlying patient differences requires further
study. 35 Although the hospitalist movement holds promise for improving end-of-life discussions,
more research is needed to determine whether this promise will be met.
End-of-Life Education for the Public
Extensive public awareness and educational programs are necessary to create a foundation for
successful end-of-life conversations in patients with advanced illness. Broadcasts, such as the
PBS-Bill Moyers special “On Our Own Terms,” may help the public appreciate the experience
of terminal illness, and the complex choices that are faced. Such presentations may encourage
viewers to discuss how they might manage a similar situation, and explore their own fears and
concerns about dying. There are no data regarding the effectiveness of public education to
improve advance care planning.
Other Locally Successful Advance Care Planning Programs
Individual programs to implement patient preferences have emerged around the country.
Limited data suggest that they may be effective, and bear further examination as to their
portability to other programs and settings and their durability over time.
• “Respecting Your Choices” Program
Gundersen Lutheran Medical Center in La Crosse, Wisconsin has worked on community-wide
programs to improve advance care planning with an initiative called “Respecting Your Choices.”
This program used patient and family education, community outreach, education for nonmedical
professionals, standard training sessions, and standard methods for documenting and
tracking advance directives. Hammes et al reported that 85% of patients in the intervention
group had written advance directives at death, executed on average 1.2 years before death. Of
these directives, 95% were in the medical record. Virtually all patients (95%) reported that the
interview process was meaningful. The patients felt that they benefited from improved
communication with loved ones and with health care providers. 36
• Dayton VA Initiative
The Dayton (Ohio) VA Medical Center aimed to increase the number of veterans who
participated in advance care planning. VA patients and their families received a patient education
booklet and a video on advance care planning. The VA also developed discussion guidelines for
providers, initiated an advance care planning clinic, and initiated a bereavement support group.
In a 12-week period, advance care planning discussions and follow-up increased from about 15%
percent of charts to almost 90%. 37
• “Let Me Decide” Program
Molloy et al examined patient satisfaction with decision making and health care costs
after systematically implementing an advance directive program in nursing homes. The “Let Me
Decide” program included educating staff in local hospitals and nursing homes, residents, and
families about advance directives and offering competent residents or next-of-kin of mentally
incompetent residents an advance directive. The researchers reported that systematic
implementation of this program reduced hospitalizations and aggressive care for nursing home
patients who did not want that level of intervention. It also reduced utilization of health care
services without affecting satisfaction or mortality. 38
562

Costs and Implementation
Estimating the cost of ascertaining and respecting patient preferences is difficult since
improvements in this area may require major changes in the structure of the health care system.
Institutional barriers, the culture of medicine, patient attitudes, time constraints physicians face
with office visits may all play a role in implementation and may inhibit change. 26 Barriers to
implementation include complacency on the part of the physician and patient, fear of political
controversy, diffused responsibility, and absence (or perverse) financial incentives for providers
and institutions. The surprising ineffectiveness of the SUPPORT intervention, which cost over
28 million dollars, demonstrates how difficult it is to make major improvements in this area.
Nevertheless, improving our ability to respect patient preferences is valuable in its own right and
may ultimately prove to be cost-effective, since some patients will choose to forego high
technology and expensive care at the end of life.
Medical care at the end of life consumes 10% to 12% of the total health care budget. An
estimated 40% of the Medicare budget is spent during the last 30 days of life. 39 Some have
posited that increased use of hospice and advance directives and lower use of high-technology
interventions for terminally ill patients will produce significant cost savings. However, the
studies on cost savings from hospice and advance directives are not definitive. The 3 randomized
trials of hospice and advance directives use show no overall savings, but the authors of a review
suggest that the studies were either too small for confidence in their negative results or their
intervention and cost accounting are flawed. 40 In the absence of a definitive study, the existing
data suggest that hospice and advance directives can save between 25% and 40% of health care
costs during the last month of life, but far less (and perhaps nothing) in the 3-12 months before
death. Although, these savings are less than most people anticipate, they do indicate that hospice
and advance directives should be encouraged because they certainly do not cost more and they
provide a means for patients to exercise their autonomy over end-of-life decisions. 40 Finally,
several of the promising interventions described above (eg, the POLST intervention), are
relatively inexpensive. For example, 500 POLST forms can be ordered from Oregon Health
Science University’s Web site 41 for less than $100, although the cost of implementing the
POLST program is unknown.
Comment
Preventing unwanted aggressive care at the end of life requires active communication
between provider and patient, and effective strategies to transfer information regarding
preferences seamlessly across care venues. The dominant strategy to improve care in this area
over the past 20 years has been the promotion of advance directives. Although the enthusiasm
for advance directives has not been matched by evidence of their effectiveness, SUPPORT and
other studies have renewed public concern and prompted providers and policy makers to
reexamine advance care planning and strive to improve it. Although we have found evidence of
several potentially promising strategies (perhaps the most promising of which is the POLST
form), the inevitability of death and the importance patients place on improving end-of-life care
point strongly to the need for further research in this area.
563

Figure 49.1. POLST Wallet Card instructions
POLST WALLET CARD INSTRUCTIONS
This is an identification wallet card for the Physician
Orders for Life-Sustaining Treatment (POLST) document.
This card is not a substitute for a completed POLST
document. It provides a summary of the POLST document
and is expected to be honored by care providers.
The POLST document and wallet card are completed by
the physician. The physician must sign both the POLST
document and the wallet card to make the wallet card valid.
Instructions continued on other side.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Physician Orders for Life-Sustaining Treatment
Name: __________________________________________
Resuscitation (Patient has no pulse and is not breathing:
Resuscitate Do Not Resuscitate (DNR)
Medical Interventions (has pulse and/or is breathing):
Antibiotics:
Comfort Measures Only Limited Interventions
Advanced Interventions Full Treatment/Resuscitation
No antibiotics except if needed for comfort
No invasive (IM/IV) antibiotics Full Treatment
Instructions continued on other side.
…continued from other side
It is recommended that the completed wallet card
564
be laminated in plastic for durability and to
prevent alteration. An existing card should be
destroyed if the POLST document is changed.
A new wallet card can be completed to match
the new physician orders.
CENTER FOR ETHICS IN HEALTH CARE
Oregon Health Sciences University
3181 SW Sam Jackson Park Rd., UHN-86
Portland, Oregon 97201-3098
(503) 494-4466
_ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Artificially Administered Fluids and Nutrition:
No feeding tube/IV fluids Full Treatment
No long term feeding tube/IV fluids
Discussed with:
Patient Health Care Representative
Court-appointed Guardian Spouse Other
I have completed the following forms:
Advance Directive Court-appointed Guardian
____________________________________ ________________
Print name of Physician Phone
____________________________________ ________________
Signature of Physician Datee
Center for Ethics in Health Care Developed in conformance with ORS 127.505 et seq.
©Center for Ethics in Health Care. Reproduced with permission.

References
1. Jonsen AR SM, Winslade WJ. Clinical Ethics: A Practical Approach to Ethical Decisions
in Clinical Medicine. Fourth ed. New York, NY: McGraw-Hill; 1998.
2. Ethics manual. Fourth edition. American College of Physicians. Ann Intern Med. 1998;
128: 576-94.
3. Teno J, Lynn J, Wenger N, Phillips RS, Murphy DP, Connors AF, Jr., et al. Advance
directives for seriously ill hospitalized patients: effectiveness with the patient selfdetermination
act and the SUPPORT intervention. SUPPORT Investigators. Study to
Understand Prognoses and Preferences for Outcomes and Risks of Treatment. J Am Geriatr
Soc. 1997; 45: 500-7.
4. anonymous. A controlled trial to improve care for seriously ill hospitalized patients. The
study to understand prognoses and preferences for outcomes and risks of treatments
(SUPPORT). The SUPPORT Principal Investigators [published erratum appears in JAMA
1996 Apr 24; 275: 1232]. JAMA 1995; 274: 1591-8.
5. Gamble ER, McDonald PJ, Lichstein PR. Knowledge, attitudes, and behavior of elderly
persons regarding living wills. Arch Intern Med. 1991; 151: 277-80.
6. Muldoon MF, Barger SD, Flory JD, Manuck SB. What are quality of life measurements
measuring? BMJ. 1998; 316: 542-5.
7. Hanson LC, Tulsky JA, Danis M. Can clinical interventions change care at the end of life?.
Ann Intern Med. 1997; 126: 381-8.
8. Morrison RS, Olson E, Mertz KR, Meier DE. The inaccessibility of advance directives on
transfer from ambulatory to acute care settings. JAMA. 1995; 274: 478-82.
9. Broadwell A, Boisaubin, EV, Dunn, JK, Engelhardt, HT. Advance directives on hospital
admission: a survey of patient attitudes. South Med J. 1993; 86: 165-8.
10. Teno JM, Licks S, Lynn J, Wenger N, Connors AF, Jr., Phillips RS, et al. Do advance
directives provide instructions that direct care? SUPPORT Investigators. Study to
Understand Prognoses and Preferences for Outcomes and Risks of Treatment. J Am Geriatr
Soc.1997; 45: 508-12.
11. Danis M, Southerland LI, Garrett JM, Smith JL, Hielema F, Pickard CG, et al. A
prospective study of advance directives for life-sustaining care. N Engl J Med. 1991; 324:
882-8.
12. Christakis N, Iwashyna TJ. Attitude and self-reported practice regarding prognostication in
a national sample of internists. Arch Intern Med. 1998; 158: 2389-95.
13. Lo B, McLeod GA, Saika G. Patient attitudes to discussing life-sustaining treatment. Arch
Intern Med. 1986; 146: 1613-5.
14. anonymous. Good care of the dying patient. Council on Scientific Affairs, American
Medical Association. JAMA. 1996; 275: 474-8.
15. Wenger NS, Phillips RS, Teno JM, Oye RK, Dawson NV, Liu H, et al. Physician
understanding of patient resuscitation preferences: insights and clinical implications. J Am
Geriatr Soc. 2000; 48: S44-51.
16. Covinsky KE, Fuller JD, Yaffe K, Johnston CB, Hamel MB, Lynn J, et al. Communication
and decision-making in seriously ill patients: findings of the SUPPORT project. The Study
to Understand Prognoses and Preferences for Outcomes and Risks of Treatments. J Am
Geriatr Soc. 2000; 48: S187-93.
17. Approaching Death: Improving Care at the End of Life. Washington DC: National
Academy Press; 1997.
565

18. Donaldson M, Field, MJ. Measuring quality of care at the End of Life. Arch Intern Med.
1998; 158: 121-28.
19. Ebell MH, Becker LA, Barry HC, Hagen M. Survival after in-hospital cardiopulmonary
resuscitation. A meta-analysis. J Gen Intern Med.1998; 13: 805-16.
20. Basta L, Plunkitt K, Shassy, R, Gamouras G. Cardiopulmonary resuscitation in the elderly:
Defining the limits of appropriateness. Am J Geriatr Cardiol. 1998; 7: 46-55.
21. Tolle SW, Tilden VP, Nelson CA, Dunn PM. A prospective study of the efficacy of the
physician order form for life-sustaining treatment. J Am Geriatr Soc. 1998; 46: 1097-102.
22. Jaret P. Leading Patients in End-of-Life Decisions. Hippocrates; 1999. p. 33-37.
23. Lee M, Brummel-Smith, K, Meyer, J, Drew, N, London, MR. Physician orders for lifesustaining
treatment (POLST): outcomes in a PACE program. Program of All-Inclusive
Care for the Elderly. J Am Geriatr Soc. 2000; 48: 1219-25.
24. Dexter P, Wolinsky, FD, Gramelspacher, GP, Zhou, ZH, Eckert, GJ, Waisburd, M,
Tierney, WM. Effectiveness of computer-generated reminders for increasing discussions
about advance directives and completion of advance directive forms. A randomized,
controlled trial. Ann Intern Med. 1998; 128: 102-10.
25. Ghusn H, Teasdale, TA, Jordan, D. Continuity of do-not-resuscitation orders between
hospital and nursing home settings. J Am Geriatr Soc. 1997; 45: 465-69.
26. Tulsky JA, Fischer GS, Rose MR, Arnold RM. Opening the black box: how do physicians
communicate about advance directives?. Ann Intern Med. 1998; 129: 441-9.
27. Rabow M, Hardie, GE, Fair, JM, McPhee, SJ. End-of-life care content in 50 textbooks
from multiple specialties. JAMA. 2000; 283: 771-8.
28. McPhee S, Rabow, MW, Pantilat, SZ, Markowitz, AJ. Finding our way- - perspectives on
care at the close of life. JAMA. 2000; 284: 2512-13.
29. Emanuel LL, von Gunten, C.F., Ferris, F.D. The Education for Physicians on End-of-Life
Care (EPEC) Curriculum; 1999.
30. Larson DG, Tobin DR. End-of-life conversations: evolving practice and theory. JAMA.
2000; 284: 1573-8.
31. Lo B. Improving care near the end of life. Why is it so hard? [editorial; comment]. JAMA.
1995; 274: 1634-6.
32. Buckman R KY. How to Break Bad News: A Guide for Health Care Professionals.
Baltimore, MD: Johns Hopkins University Press; 1992.
33. Lo B, Quill T, Tulsky J. Discussing palliative care with patients. ACP-ASIM End-of-Life
Care Consensus Panel. American College of Physicians-American Society of Internal
Medicine. Ann Intern Med. 1999; 130: 744-9.
34. Pantilat SZ, Alpers A, Wachter RM. A new doctor in the house: ethical issues in hospitalist
systems. JAMA. 1999; 282: 171-4.
35. Auerbach AD, Pantilat SZ, Wachter RM, Goldman L. Processes and outcomes of end-oflife
care in a voluntary hospitalist model. J Gen Intern Med. 2001; 16: 115.
36. Hammes BJ, Rooney BL. Death and end-of-life planning in one midwestern community.
Arch Intern Med. 1998; 158: 383-90.
37. Lynn J SJ, Kabcenell A. Beyond the Living Will: Advance Care Planning for All Stages of
Health and Disease. Improving Care for the End of Life: A Sourcebook for Health Care
Managers and Clinicians: Oxford University Press; 2000. p. 73-90.
38. Molloy DW, Guyatt GH, Russo R, Goeree R, O'Brien BJ, Bedard M, et al. Systematic
implementation of an advance directive program in nursing homes: a randomized
controlled trial. JAMA. 2000; 283: 1437-44.
566

39. Kurent J. Death and Dying in America: The need to improve End-of-Life Care. Carolina
Healthcare Business 2000:16-19.
40. Emanuel EJ. Cost savings at the end of life. What do the data show? JAMA. 1996; 275:
1907-14.
41. Center for Ethics in Health Care, Oregon Health Sciences University,
http://www.ohsu.edu/ethics/polst.htm, last accessed May 8, 2001.
567

568

Chapter 50. Other Practices Related to Patient Participation
Laura T. Pizzi, PharmD
Neil I. Goldfarb
David B. Nash, MD, MBA
Thomas Jefferson University School of Medicine
and Office of Health Policy & Clinical Outcomes
Background
A number of practices and resources aim to facilitate the role of patients as their own
safety advocates. These practices are not intended to shift the burden of monitoring medical error
to patients. Rather, they encourage patients to share responsibility for their own safety. Although
these types of interventions hold promise for enhanced patient safety, there is yet insufficient
evidence of their effectiveness. Therefore, this chapter is a brief, general survey of practices
related to patient participation; there are few practices that have been studied with sufficient
rigor to merit a full evidence-based review. This chapter explicitly excludes consumer report
cards, since such tools presently are more relevant to health care quality than patient safety. 1,2
There is a substantial literature on the patient’s role in quality improvement related to specific
diseases - eg, self-management and general education for patients with certain chronic diseases
such as asthma, 3-5 diabetes, 6-8 and rheumatoid arthritis, 9-12 as well preoperative educational and
preparation programs for patients undergoing cardiac surgery. 13 This literature was not reviewed
in detail, both because it falls outside our definition of patient safety practices (Chapter 1) by
virtue of its disease specificity, and because the volume of material was overwhelming given the
time allocated for the production of this Report. There are obvious additional opportunities to
promote patient involvement in helping protect their own safety drawn from the disease-specific
experiences of the past, and this should be the subject of further exposition and analysis.
Patient Education Materials Regarding Patient Safety
Books, Web sites and consumer group publications abound with health care and medical
information for patients. 14 The goal of these resources is to enable consumers to arm themselves
with the knowledge to protect themselves. Health care providers may wish to distribute such
materials to patients to alert them of the possible problem of medical error, and encourage those
that would like to take appropriate action.
The Agency for Healthcare Research and Quality 15 produces a 5-page “Patient Fact
Sheet” on preventing medical errors. This fact sheet educates patients on the problem of medical
error, and provides 20 tips patients may follow to avoid medical error, ranging from properly
measuring liquid medications to ensuring health care employees have washed their hands.
Proprietary educational materials have also been developed. For instance, DoctorQuality,
Inc., a quality management company that provides products and services to health care
consumers, purchasers, and providers, has developed online and offline tools that providers and
patients can use to improve care, 16 including patient safety workbooks and quality guides for a
variety of diagnoses and surgical procedures. The books describe the key events that patients
should anticipate at each step of diagnosis and treatment, identify high-risk points in the
treatment plan where mistakes are more likely occur, 14 and provide tips as to how to avert
common errors.
Patients may also find resources in the popular literature. In Lerner’s Consumer Guide to
Health Care, the authors coach readers on questions to ask their physicians and ways to avoid
569

medical mistakes. 17 Dr. Robert Arnot's book, The Best Medicine, 18 educates patients about
specific procedures (eg, coronary artery bypass surgery, cesarean section, hysterectomy, and
carotid endarterectomy). Potential complications of each of these procedures are described, and
volumes, average lengths of stay, and complication rates of major hospitals are presented.
Health information on the Web has increased patients’ desire for medical information and
raised significant issues regarding patient safety and the manner in which patients approach their
doctors for information. 19 A recent study revealed that many physicians believe that Web
resources can distance patients from physicians and have an adverse effect on patient safety.
Specifically, there is concern that patients can receive and believe misinformation, or read of
treatments and procedures unfamiliar to physicians, and in both instances lower their trust in
physician recommendations. 20 Other physicians see the Web as a positive development in patient
safety because when patients approach their doctors prepared with questions, office visits run
more smoothly, and the physician’s counsel may be better received. 20
Similar issues surround the topic of direct-to-consumer (DTC) marketing by
pharmaceutical companies. 21 In 1997 the Food and Drug Administration relaxed restrictions on
television and advertising for prescription medication. Drug companies responded with an
explosion of marketing in all forms of media. DTC advertising may stimulate discussion
between patients and their doctors about treatment options, but it also drives patients to demand
newer and costlier medications, when less expensive treatments might be effective. 22 When
doctors resist, 46% of patients try to persuade them to change the original recommendation and
another 24% attempt to obtain the requested drugs from another physician. 22 These sort of
interactions erode the physician-patient relationship and may jeopardize safety by promoting
polypharmacy.
Practices to Improve Non-compliance
Compliance with medical advice is widely discussed in the literature and non-compliance
with treatment may result in adverse drug events. 23 The frequency of non-compliance is higher
than many health care professionals realize. 24 Non-compliance may arise from
misunderstandings regarding instructions for drug use, self-care, or other factors. In addition,
some of these misunderstandings may arise from remediable factors, such as language barriers 25
or low health literacy. 26,27 Simple solutions, such as using a trained interpreter instead of a family
member or untrained volunteer, and providing self-care and other literature in multiple languages
and bilingual versions may improve patient understanding. Other interventions, such as patient
education publications, have been proposed to reduce adverse drug events due to noncompliance.
28
Access to Medical Records
Although patient access to their own medical records is regulated in some states, these
statutes differ across the United States. 29 Some states mandate certain levels of access for
patients; others limit access or allow the provider to deem access appropriate or inappropriate.
Other countries, such as Britain, have passed legislation requiring that providers allow patients to
have complete access to their medical records. 30 Some argue that access to medical records may
encourage patients to take a more active role in their own health care, allow patients to become
better informed about their care, and increase rapport. Others argue that staff may modify
medical records due to concerns about offending the patient, and will be diverted by the time
needed to explain information contained in the records. Finally, still others express concern that
patients may be unnecessarily worried by what they read. 30 No studies in the United States have
570

analyzed these competing views, and therefore it is not clear whether cultural norms reported in
studies from other countries are applicable here, and whether allowing patients to review their
own charts will have the intended effect of reducing errors.
Comment
With the growing level of consumerism in health care, 31 patients may wish to take a more
active role in reducing their chance of experiencing a medical error. However, the research
regarding the ways in which providers can facilitate this role for patients who desire it is lacking.
More research is needed on the efficacy of these interventions regarding medical error reduction
and on patients’ willingness and ability to use them.
References
1. Longo DR, Land G, Schramm W, Fraas J, Hoskins B, Howell V. Consumer reports in
health care. Do they make a difference in patient care? JAMA. 1997; 278: 1579-1584.
2. Marshall MN, Shekelle PG, Leatherman S, Brook RH. The public release of performance
data: what do we expect to gain? A review of the evidence JAMA. 2000; 283: 1866-1874.
3. Lahdensuo A, Haahtela T, Herrala J, Kava T, Kiviranta K, Kuusisto P, et al. Randomised
comparison of guided self management and traditional treatment of asthma over one year.
BMJ. 1996; 312: 748-752.
4. Gibson PG, Coughlan J, Wilson AJ, Abramson M, Bauman A, Hensley MJ, et al. Selfmanagement
education and regular practitioner review for adults with asthma. In: The
Cochrane Library, Issue 2, 2001. Oxford: Update Software.
5. Gibson PG, Coughlan J, Wilson AJ, Hensley MJ, Abramson M, Bauman A, et al. Limited
(information only) patient education programs for adults with asthma. In: The Cochrane
Library. 2001. Oxford: Update Software.
6. Fain JA, Nettles A, Funnell MM, Charron D. Diabetes patient education research: an
integrative literature review. Diabetes Educ. 1999; 25: 7-15.
7. Norris SL, Engelgau MM, Narayan KM. Effectiveness of self-management training in type
2 diabetes: a systematic review of randomized controlled trials. Diabetes Care. 2001; 24:
561-587.
8. Hampson SE, Skinner TC, Hart J, Storey L, Gage H, Foxcroft D, et al. Effects of
educational and psychosocial interventions for adolescents with diabetes mellitus: a
systematic review. Health Technol Assess. 2001; 5: 1-79.
9. Superio-Cabuslay E, Ward MM, Lorig KR. Patient education interventions in osteoarthritis
and rheumatoid arthritis: a meta-analytic comparison with nonsteroidal antiinflammatory
drug treatment. Arthritis Care Res. 1996; 9: 292-301.
10. Brus H, van de Laar M, Taal E, Rasker J, Wiegman O. Compliance in rheumatoid arthritis
and the role of formal patient education. Semin Arthritis Rheum. 1997; 26: 702-710.
11. Brus HL, van de Laar MA, Taal E, Rasker JJ, Wiegman O. Effects of patient education on
compliance with basic treatment regimens and health in recent onset active rheumatoid
arthritis. Ann Rheum Dis. 1998; 57: 146-151.
12. Barlow JH, Turner AP, Wright CC. A randomized controlled study of the Arthritis Self-
Management Programme in the UK. Health Educ Res. 2000; 15: 665-680.
13. Arthur HM, Daniels C, McKelvie R, Hirsh J, Rush B. Effect of a preoperative intervention
on preoperative and postoperative outcomes in low-risk patients awaiting elective coronary
artery bypass graft surgery. A randomized, controlled trial. Ann Intern Med. 2000; 133:
253-262.
571

14. Robinson JL, Nash DB. Consumers' role in patient safety. QRC Advis. 2000; 17: 1-3.
15. Patient Fact Sheet. 20 Tips to Help Prevent Medical Errors (AHRQ Publication No. 00-
PO38) Agency for Healthcare Research and Quality. Available at:
http://www.ahrq.gov/consumer/20tips.htm. Accessed June 25, 2001.
16. DoctorQuality corporate website DoctorQuality.com, Inc. Available at:
http://www.doctorquality.com/www/about.asp. Accessed June 25, 2001.
17. Lerner P, Lerner J. Lerner's consumer guide to health care: how to get the best health care
for less. Seattle, WA: Lerner Communications; 2001.
18. Arnot RB. The Best Medicine: How to Choose the Top Doctors, the Top Hospitals, and the
Top Treatments. Reading, MA: Addison-Wesley Publishing Company; 1993.
19. Fotch E. The effect of the world wide web on you and your patients. Ophthalmol Clin
North Am. 2000; 13: 261-269.
20. Bard M. How will the web affect the physician-patient relationship? Interview by Marilynn
Larkin. Lancet. 2000; 356: 1777.
21. Wilkes MS, Bell RA, Kravitz RL. Direct-to-consumer prescription drug advertising: trends,
impact, and implications. Health Aff (Millwood). 2000; 19: 110-128.
22. Brown AB. The direct-to-consumer advertising dilemma. Patient Care 2001; 35:22-33.
2001; 35: 22-33.
23. Holt GA, Dorcheus L, Hall EL, Beck D, Ellis E, Hough J. Patient interpretation of label
instructions. Am Pharm. 1992; NS32: 58-62.
24. Fletcher RH. Patient compliance with therapeutic advice: a modern view. Mt Sinai J Med.
1989; 56: 453-458.
25. Woloshin S, Bickell NA, Schwartz LM, Gany F, Welch HG. Language barriers in medicine
in the United States. JAMA. 1995; 273: 724-728.
26. Williams MV, Parker RM, Baker DW, Parikh NS, Pitkin K, Coates WC, et al. Inadequate
functional health literacy among patients at two public hospitals. JAMA. 1995; 274: 1677-
1682.
27. Gazmararian JA, Baker DW, Williams MV, Parker RM, Scott TL, Green DC, et al. Health
literacy among Medicare enrollees in a managed care organization. JAMA. 1999; 281: 545-
551.
28. Medication safety issue brief. Asking consumers for help. Part 3. Hosp Health Netw. 2001;
75: suppl 2 p. following 56.
29. Patient Confidentiality American Medical Association. Available at: http://www.amaassn.org/ama/pub/category/4610.html.
Accessed June 25, 2001.
30. Grange A, Renvoize E, Pinder J. Patients' rights to access their healthcare records. Nurs
Stand. 1998; 13: 41-42.
31. Nash DB, Manfredi MP, Bozarth B, Howell S. Connecting with the new healthcare
consumer. Gaithersburg, MD: Aspen Publishers, Inc.; 2001.
572

PART IV. PROMOTING AND IMPLEMENTING SAFETY PRACTICES
Chapter 51. Practice Guidelines
Chapter 52. Critical Pathways
Chapter 53. Clinical Decision Support Systems
Chapter 54. Educational Techniques Used in Changing Provider Behavior
Chapter 55. Legislation, Accreditation, and Market-Driven and Other Approaches to
Improving Patient Safety
573

574

Chapter 51. Practice Guidelines
Robert Trowbridge, MD
University of California, San Francisco School of Medicine
Scott Weingarten, MD, MPH
University of California, Los Angeles School of Medicine
Background
Practice guidelines are among the most widely employed methods of modifying
physician behavior. Defined as “systematically developed statements to assist practitioner and
patient decisions about appropriate health care for specific clinical conditions,” 1 guidelines may
affect both the process and the outcome of care. Although guideline development and
implementation have traditionally focused on ensuring a perceived standard of care, increasing
emphasis has been placed on patient outcomes and patient safety. In this regard, thousands of
guidelines have been promulgated on a great variety of clinical topics ranging from the
prevention of falls in the elderly to the proper use of bone marrow transplantation. 2,3
Practice Description
Guidelines vary greatly in terms of both method of development and format. Some
consist of relatively straightforward statements that advocate a particular clinical practice,
whereas others represent a series of complex algorithms that require the input of multiple clinical
variables. Many guidelines are developed by specialty and advocacy organizations with attention
paid to rigorously conducted systematic reviews. 4 Others may simply reflect a local standard of
care.
Guidelines should only be considered a potential source of information and are effective
at modifying physician behavior only when coupled with appropriate implementation strategies.
Guidelines disseminated using a multifaceted approach that provides for peer influence and
management support, for example, are more likely to be successful in influencing provider
behavior than strategies that depend solely on the passive dissemination of printed materials. 5
Prevalence and Severity of Target Safety Problem/Opportunities for Impact
Practice guidelines have the potential to greatly impact patient safety as they may
facilitate widespread dissemination of practices that effectively reduce medical errors. It is well
established in other medical fields that known tenets of effective health care are not practiced on
a universal basis despite overwhelming evidence supporting their use. According to an audit of
Medicaid charts in Connecticut, for example, only 50% of patients presenting with an acute
myocardial infarction received aspirin and beta-blockers on the day of admission despite
substantial evidence that these practices reduce mortality. 6 A second report estimated that 3500
infarctions would be averted and 4300 lives saved annually if beta-blockers were appropriately
prescribed in patients with coronary artery disease. 7 Guidelines could help rectify similar
shortcomings in the field of patient safety with corresponding reductions in medical errors.
575

Study Design
There are no well-designed studies that specifically evaluate the effect of practice
guidelines on patient safety. However, research regarding their effectiveness in modifying
physician behavior and improving patient outcomes is plentiful. The most comprehensive review
evaluating general utility of guidelines involved an extensive search of MEDLINE and other
electronic databases, the gray literature and the bibliographies of pertinent articles. Its analysis
included 59 reports consisting of both controlled time series and before-after studies in addition
to randomized trials. 8
A second systematic review, limited to computer-based guidelines, reported the results of
a search of several electronic databases (MEDLINE and CINAHL) complemented by a limited
bibliography search. It included 25 studies detailing the use of 20 guideline systems. In this
group, there were 10 time series studies (all without external controls) and 10 controlled trials, 9
of which were randomized. 9
A final systematic review, limited to the primary care setting, also searched MEDLINE
and several other electronic databases and included a limited bibliography search. 10 This review
included only randomized trials that reported clinical outcomes associated with the treatment of
patients with established diagnoses and identified 13 studies for inclusion. In keeping with these
criteria, trials of the use of guidelines to promote preventive health care and proper diagnostic
evaluations were excluded.
Several studies not represented in the above systematic reviews also provide valuable
information on the effectiveness of practice guidelines. Two multicenter studies, for example,
evaluated the impact of guidelines on the treatment of patients admitted for pneumonia and
certain surgical procedures, respectively. Both of these were prospective before-after trials. 11,12 A
prospective controlled trial of the implementation of a diabetes guideline using problem-based
learning was also completed, 13 as were 2 randomized controlled trials of the local
implementation of nationally developed guidelines. 14,15 Other recently completed works include
an evaluation of a guideline for the outpatient treatment of cystitis with concurrent and historical
controls, and a prospective before- after investigation of a guideline for the treatment of upper
gastrointestinal hemorrhage. 16,17
Study Outcomes
Very few of the reviewed studies report outcomes specifically linked to patient safety.
Those that do (such as guidelines to prevent falls in the elderly (Chapter 26) and to promote hand
washing (Chapter 12)) are of less than robust design. 2,18,23 A few of the studies included in the
systematic reviews do report the effect of guidelines on surgical site infections and vaccination
rates, but results are reported in terms of the process of care rather than the outcome of care. The
systematic review from Worrall et al is an exception, as it only analyzed studies that reported
direct clinical variables. 10 Among the other studies, however, the most commonly reported
variables were length of stay, rates of provider adherence to guidelines, complication rates and
the rates of the appropriate use of diagnostic testing.
576

Evidence for Effectiveness of Practice
Despite the inherent methodologic problems of many of the cited studies, there is
substantial evidence that practice guidelines may be effective in influencing provider behavior
and patient outcomes. In the seminal review from Grimshaw and Russell, 9 of 11 studies
reporting clinical outcomes noted some degree of improvement while the process of care was
found to improve in 55 of 59 studies. 8 This review did include several trials with marginal study
design, but the authors argued for the inclusion of the time series and before-after studies on the
premise that the magnitude of effect seen in many of the studies overwhelmed any potential bias
that may have been attributable to study design. They additionally noted that the randomized
trial may not represent the best method for evaluating practice guidelines.
The systematic review investigating the utility of computer-based guidelines is similarly
encouraging. 9 Fourteen of 18 studies found guidelines increased provider adherence to the tenets
of care promoted by the guidelines. Of the 8 studies that evaluated patient outcomes, 3 found
improvements. These results mirror those of Grimshaw and Russell and reinforce the supposition
that guidelines, at least in the studies completed to date, are more effective at influencing the
process rather than the outcome of care.
The systematic review of controlled trials reporting clinical outcomes in the primary care
setting is significantly less optimistic. 10 Only 5 of the 13 trials analyzed showed improvements in
the defined clinical outcomes. None did so across all of the study groups or for an extended
period of time. These results cast doubt on the utility of guidelines, particularly since this paper
only included rigorously conducted trials reporting clinical outcomes, in contrast to the larger
review by Grimshaw and Russell. However the authors correctly assert that many of these
studies likely used guidelines that were not evidence-based and which may not have been
sensitive enough to discern small improvements in clinical outcomes. In addition, it is likely that
improvements in the process of care represents a reasonable surrogate endpoint for clinical
outcomes given the size of the studies that would need to be conducted to show a beneficial
outcomes effect of guidelines. To exclude all studies using this surrogate endpoint may be an
extreme and unnecessary measure.
The studies of guidelines that have been published since the completion of the above
systematic reviews are similarly conflicting. The multicenter study of the pneumonia guideline,
for example, showed no effect on the length of stay or complication rates whereas the study of
the postoperative surgical guidelines showed a significant decrease in the length of stay in 2 of
the 3 groups. 11,12 Additionally, studies of the implementation of nationally developed guidelines
yielded conflicting results, perhaps as a function of the implementation strategies used in each
study. The study of the Agency for Healthcare Policy and Research (AHCPR, now the Agency
for Healthcare Research and Quality, AHRQ) guideline for cessation of smoking in pregnant
woman, which was a rigorously conducted randomized controlled trial, used extensive
implementation strategies and showed a marked and statistically significant increase in the
smoking cessation rate in the intervention group. 15 In contrast, the trial using continuous quality
improvement measures and academic detailing to promote the AHCPR depression guideline
showed little effect. 14 Yet several other studies, all of reasonable design and well-controlled,
demonstrated improvements in the process of care 13,16,17 with one also reporting a marked
clinical benefit in improving glycemic control in diabetics. 13
577

Potential for Harm
It has been theorized that practice guidelines, if improperly developed or implemented,
could actually be detrimental to the process of care or worsen clinical outcomes. A study of
guidelines in the use of neurodiagnostic testing in patients with low back pain found the tests
were more frequently utilized in an improper fashion than at baseline when clinicians were given
a set of guidelines that were relatively narrow in focus. 19 Another study of patients treated for
congestive heart failure found the guidelines actually increased the length of stay beyond that
which was clinically necessary. 20 Although neither of these studies revealed worsened patient
outcomes due to the guidelines, that potential certainly exists. In addition, the promulgation of
guidelines with imprudent advice could also result in widespread harm to patients.
Costs and Implementation
Developing and implementing practice guidelines is expensive. When developed for a
complex clinical problem by a national organization, for example, they consume tremendous
resources, often tens of thousands of dollars. 21 In addition, some of the more successful
implementation strategies, such as academic detailing (Chapter 54), also require a substantial
measure of effort and financial outlay.
The manner in which practice guidelines are implemented is at least as important as the
content of the guidelines themselves. Several systematic reviews have investigated both the
strategies that are associated with successful institution of guidelines and the specific barriers to
implementation. 5, 22, 23 These reviews concluded that the deterrents vary by practice location and
that strategies to circumvent these barriers must be devised on an individual and local basis. In
general, however, the number and intensity of the implementation strategies employed generally
corresponds with the ultimate success of the guideline.
Comment
There is convincing but by no means overwhelming evidence that practice guidelines are
effective in positively influencing the process and, to a lesser extent, the outcome of care. Many
of the completed studies are plagued by methodologic shortcomings that reflect the difficulty
inherent in studying the impact of guidelines. Although evidence specific to the use of guidelines
in patient safety is scanty, they are likely to be as effective in this area as any other. It thus
appears that well-constructed guidelines could play a significant role in ensuring patient safety
and reducing medical errors. The effectiveness of guidelines, however, is dependent on many
factors outside of their content. In particular, specific attention must be focused on utilizing
appropriate implementation strategies if the full potential of guidelines is to be realized.
References
1. Field MJ LK, eds. Clinical Practice Guidelines: Directions for a new program.
Washington, DC: National Academy Press; 1990.
2. Baraff LJ, Lee TJ, Kader S, Della Penna R. Effect of a practice guideline for emergency
department care of falls in elder patients on subsequent falls and hospitalizations for
injuries. Acad Emerg Med. 1999; 6:1224-31.
3. Rowe JM, Ciobanu N, Ascensao J, Stadtmauer EA, Weiner RS, Schenkein DP, et al.
Recommended guidelines for the management of autologous and allogeneic bone marrow
transplantation. A report from the Eastern Cooperative Oncology Group (ECOG). Ann
Intern Med. 1994; 120: 143-58.
578

4. Weingarten S. Using practice guideline compendiums to provide better preventive care.
Ann Intern Med. 1999; 130: 454-8.
5. Wensing M, van der Weijden T, Grol R. Implementing guidelines and innovations in
general practice: which interventions are effective? Br J Gen Pract. 1998; 48: 991-7.
6. Holmboe ES, Meehan TP, Radford MJ, Wang Y, Marciniak TA, Krumholz HM. Use of
critical pathways to improve the care of patients with acute myocardial infarction. Am J
Med. 1999; 107: 324-31.
7. Phillips KA, Shlipak MG, Coxson P, Goldman L, Heidenreich PA, Weinstein MC, et al.
Underuse of beta-blockers following myocardial infarction. JAMA. 2001; 285: 1013.
8. Grimshaw JM, Russell IT. Effect of clinical guidelines on medical practice: a systematic
review of rigorous evaluations. Lancet. 1993; 342: 1317-22.
9. Shiffman RN, Liaw Y, Brandt CA, Corb GJ. Computer-based guideline implementation
systems: a systematic review of functionality and effectiveness. J Am Med Inform Assoc.
1999; 6: 104-14.
10. Worrall G, Chaulk P, Freake D. The effects of clinical practice guidelines on patient
outcomes in primary care: a systematic review. CMAJ. 1997; 156: 1705-12.
11. Weingarten S, Riedinger MS, Sandhu M, Bowers C, Ellrodt AG, Nunn C, et al. Can
practice guidelines safely reduce hospital length of stay? Results from a multicenter
interventional study. Am J Med. 1998; 105: 33-40.
12. Rhew DC, Riedinger MS, Sandhu M, Bowers C, Greengold N, Weingarten SR. A
prospective, multicenter study of a pneumonia practice guideline. Chest. 1998; 114: 115-9.
13. Benjamin EM, Schneider MS, Hinchey KT. Implementing practice guidelines for diabetes
care using problem-based learning. A prospective controlled trial using firm systems.
Diabetes Care. 1999; 22: 1672-8.
14. Brown JB, Shye D, McFarland BH, Nichols GA, Mullooly JP, Johnson RE. Controlled
trials of CQI and academic detailing to implement a clinical practice guideline for
depression. Jt Comm J Qual Improv. 2000; 26: 39-54.
15. Windsor RA, Woodby LL, Miller TM, Hardin JM, Crawford MA, DiClemente CC.
Effectiveness of Agency for Health Care Policy and Research clinical practice guideline
and patient education methods for pregnant smokers in medicaid maternity care. Am J
Obstet Gynecol. 2000; 182: 68-75.
16. Saint S, Scholes D, Fihn SD, Farrell RG, Stamm WE. The effectiveness of a clinical
practice guideline for the management of presumed uncomplicated urinary tract infection
in women. Am J Med. 1999; 106: 636-41.
17. Hay JA ML, Weingarten SR, Ellrodt AG. Prospective evaluation of a clinical guideline
recommending hospital length of stay in upper gastrointestinal tract hemorrhage. JAMA.
1997: 2151-6.
18. Baraff LJ, Lee TJ, Kader S, Della Penna R. Effect of a practice guideline on the process of
emergency department care of falls in elder patients. Acad Emerg Med. 1999; 6: 1216-23.
19. Shekelle PG, Kravitz RL, Beart J, Marger M, Wang M, Lee M. Are nonspecific practice
guidelines potentially harmful? A randomized comparison of the effect of nonspecific
versus specific guidelines on physician decision making. Health Serv Res. 2000; 34: 1429-
48.
20. Weingarten S, Riedinger M, Conner L, Johnson B, Ellrodt AG. Reducing lengths of stay in
the coronary care unit with a practice guideline for patients with congestive heart failure.
Insights from a controlled clinical trial. Med Care. 1994; 32: 1232-43.
579

21. Smith WR. Evidence for the effectiveness of techniques to change physician behavior.
Chest. 2000; 118: 8S-17S.
22. Davis DA, Taylor-Vaisey A. Translating guidelines into practice. A systematic review of
theoretic concepts, practical experience and research evidence in the adoption of clinical
practice guidelines. CMAJ. 1997; 157: 408-16.
23. Cabana MD, Rand CS, Powe NR, Wu AW, Wilson MH, Abboud PA, et al. Why don't
physicians follow clinical practice guidelines? A framework for improvement. JAMA.
1999; 282: 1458-65.
580

Chapter 52. Critical Pathways
Robert Trowbridge, MD
University of California, San Francisco School of Medicine
Scott Weingarten, MD, MPH
University of California, Los Angeles School of Medicine
Background
Burgeoning concerns regarding patient safety, variable health care quality and increasing
health care costs have led to the introduction of clinical management tools that have their origins
outside of the traditional health care sector. Primary among these innovations has been the
implementation of critical pathways, administrative models that streamline work and production
processes. 1 Critical pathways have been utilized extensively in several different business sectors
including the construction and automotives industries. 2-4 It is theorized that the adaptation of
pathways to health care, particularly inpatient care, may help ensure the delivery of quality care
and decrease the occurrence of medical errors.
Practice Description
Although closely related to clinical practice guidelines (Chapter 51), pathways more
directly target the specific process and sequence of care, frequently plotting out the expected
course of an illness or procedure with associated prompts for appropriate interventions. Also
known as clinical pathways and care maps, pathways are generally multidisciplinary by design
and may incorporate the responsibilities of physicians and nurses with those of ancillary medical
providers including pharmacists, physical therapists and social workers. 5 They are regularly
intercalated into the point-of-care and may, in some cases, incorporate or even replace traditional
chart documentation. In addition, pathways are often evidence-based and may even be integrated
with locally or nationally developed clinical practice guidelines. Most pathways, however, are
locally developed and are most frequently implemented at the level of the hospital or medical
center as part of a cost-containment or quality assurance initiative.
Prevalence and Severity of the Problem/Opportunities for Impact
It is well established that the methods currently used to disseminate medical information
are both cumbersome and inefficient. Even medical advances that are well entrenched in the
literature and of unquestionable value are not routinely or universally implemented. A study of
patients treated for myocardial infarction in Connecticut, for example, revealed only 50% of
patients received the most basic beneficial treatments (aspirin and beta-blockers) at the time of
admission. 6 A second report suggested that 3500 myocardial infarctions would be averted and
4300 lives would be saved annually if all eligible patients with coronary artery disease were
prescribed beta-blockers. 7 Critical pathways, if able to beneficially alter health care provider
performance and ensure that effective patient safety strategies were practiced on a widespread
basis, could be powerful agents in the prevention of medical errors, in addition to any beneficial
impact they might have on overall health care quality.
581

Study Design
There is a dearth of well-designed studies analyzing the extent to which critical pathways
change physician behavior and patient outcomes. Even fewer relate specifically to the topic of
patient safety. There are no systematic reviews of pathways and most of the published work
describes the non-randomized implementation of pathways in which it is often difficult to
differentiate effects of the pathway from secular trends. The vast majority of these studies
describe the implementation of a pathway for a specific surgical procedure and use historical
controls with a retrospective before-after study design. 8-19
Four randomized controlled trials investigated the impact of the implementation of
pathways. 20-23 There are also several studies with non-randomized concurrent controls with
enrollment in the intervention arms being completed at the request of the attending physician. 24,25
The latter group of studies also included historical control groups. There is one prospective
before-after trial in which the control group actually shifted to receive the intervention after a
washout period. 26 Finally one study used historical controls and concurrent controls from other
hospitals in the region. 27 Table 52.1 describes the salient features of these papers.
Study Outcomes
The great majority of the cited studies reported at least one clinical outcome (Level 1)
with the most commonly reported variable being diagnosis-related complications. Very few of
the studies reported more definitive endpoints, such as mortality. Most of the included studies
reported surrogate clinical end-points (Level 2) as the major study outcome variables, usually
length of stay and re-admission rates. The relative utilization of certain interventions including
medications, laboratory tests and radiology studies was also commonly reported. In addition,
most of the studies also included an analysis of the changes in costs associated with instituting
the pathway. Some of the studied pathways did include other recommendations for interventions
germane to the field of patient safety, including the use of indwelling urinary catheters (Chapter
15) and prophylactic preoperative antibiotics (Chapter 20.1), but these outcomes were rarely
reported.
Evidence for Effectiveness of the Practice
Several of the randomized controlled trials provide at least some evidence that critical
pathways can be effective in influencing health care provider behavior. One study evaluated the
effectiveness of a pathway for the treatment of asthma in children admitted to a non-ICU
hospital setting. 22 Those patients treated under the pathway, when compared to the control group
undergoing “usual care,” had a shorter length of stay as well as decreased hospital charges and
medication usage, but no change in complication rates. Although the results are somewhat
promising, the study was likely skewed by a significant Hawthorne effect, as patients treated
under the pathway were placed on a separate clinical ward than those undergoing usual care
(although the same physicians cared for both groups of patients). This study found no impact on
the rate of complications, which provides little encouragement regarding the ability of pathways
to reduce medical errors and enhance patient safety.
A second randomized trial investigated the utility of a pathway for the treatment of
patients undergoing knee and hip replacement at an academic medical center in Australia. 21
Implementation of the pathway was followed by significant decreases in the length of stay and
shorter times to patient mobilization and ambulation. More importantly, it showed a decrease in
the incidence of medical complications, including readmission rates (although this change did
582

not reach statistical significance). The study, although fairly small at 163 total patients,
represents some of the most convincing evidence that pathways may be effective in decreasing
complications.
A third randomized trial, performed by the same Australian group, evaluated the effect of
a pathway for the treatment of patients with hip fracture. 20 Implementation of this pathway,
which provided recommendations regarding medications, laboratory and radiology testing and
discharge planning, resulted in a significantly shorter lengths of stay without any concomitant
change in complication rates. The study was rigorously conducted and complication rates were
meticulously documented, but no information regarding the effort needed from clinicians to
comply with the pathway was presented.
A final trial used cluster randomization to investigate the effectiveness of a critical
pathway for the treatment of community-acquired pneumonia in 19 Canadian hospitals. 23 The
pathway, which was initiated upon presentation of the patient to the emergency department,
included recommendations for the use of a specific antibiotic and provided a clinical prediction
tool to aid in decisions regarding hospital admission. Following admission, a study nurse also
regularly placed guideline-based recommendations in the chart regarding changing to oral
antibiotic therapy and discharge planning. The pathway showed impressive results in terms of
cost containment, with a shorter length of stay and a smaller percentage of inappropriate
admissions. However, there were no significant changes in the clinical parameters measured,
including mortality and complication rates. In addition, it is difficult to dissect the effect of the
pathway from that of the other elements of the multifaceted intervention.
The second strata of studies were less encouraging. A prospective before-after evaluation
of the implementation of a pathway for the administration of supplemental oxygen for
hospitalized patients showed the pathway was associated with markedly elevated costs without
any significant clinical benefit. 26 This study, which used an intensive 3-tiered strategy including
an oxygen order form, the posting of the pathway in patient rooms and a research nurse
providing immediate audit and feedback, doubled the cost of providing supplemental oxygen.
Although it did change some of the physician prescription practices, this alteration in practice
resulted in no appreciable clinical effect.
At least 2 other studies compared an intervention group with a non-randomized
concurrent control group (thus introducing the possibility of selection bias) as well as to
historical controls. In a study evaluating the impact of a critical pathway on the treatment of
patients undergoing neck dissection, the length of stay and total costs were significantly lower
for the pathway group when compared with those of the historical controls. However, the
differences disappeared when the concurrent control group was used. 25 In the other, a study of a
pathway for the treatment of asthma exacerbations, the pathway resulted in improvements in
resource utilization that were significant compared to both historical and concurrent controls. 24
However no changes were noted between the groups in readmission rates or medical outcomes.
Finally, there are a great number of observational before-after studies of pathways, the
vast majority of which relate to specific surgical procedures. 8-19 Few of these studies account for
secular trends and their capacity for judging the effectiveness of critical pathways is limited, at
best. In addition, a recent evaluation of pathways for patients undergoing a variety of surgical
procedures compared the intervention group to both historical controls and to patients from
similar hospitals in the same region in an attempt to correct for such secular trends. Although
implementation of the pathways resulted in significant decreases in length of stay when
compared to the historical controls, these differences disappeared when the pathway groups were
compared to the concurrent control groups from the other hospitals. 27 An additional study of the
583

effect of a pathway on the treatment of acute myocardial infarction demonstrated that the
improvements seen in the intervention group were likely secondary to secular trends and not an
independent effect of the pathway. 6 These powerful results cast a great deal of doubt over those
studies that demonstrated effectiveness of the pathways in comparison to historical controls
alone.
Potential for Harm
There are theoretical concerns that pathways may result in adverse patient outcomes as a
result of shortened length of stay and a dependence on “cookbook medicine,” although there is
little support for this in the literature.
Costs and Implementation
Although many studies report cost savings associated with instituting pathways, very few
detail the costs of developing and implementing them. One report attempted to put a dollar figure
on the development of a pathway for the treatment of patients undergoing knee replacement
surgery; however the reported estimate ($21,000) did not account for the time staff physicians
spent on the project. 14 The expense of developing critical pathways in terms of physician time
commitment and actual financial outlay is unknown. In addition, most pathways are developed
on the local level and require a great deal of initiative and expertise on the part of the hospital or
medical center. Whether most centers have access to these resources is unclear.
Physicians and other providers are also not universally welcoming of critical pathways.
They are considered intrusive by some providers and evidence of “cookbook medicine” by
others. The developers of pathways must be careful to allow for clinical judgment and flexibility
in the pathways or they are likely to be ignored or applied too rigidly.
Comment
There is conflicting evidence regarding the efficacy of critical pathways as a method to
modify health care provider behavior and a means of implementing patient safety initiatives.
Although a few studies suggest they may impact physician practice and, to a lesser extent,
complication rates and other clinical outcomes, the data are inconsistent and more studies are
needed. Additionally it is unclear whether the costly development and implementation of
pathways represents an appropriate use of limited health care resources. Finally, there is very
little information on the application of pathways to patient safety.
584

Table 52.1. Key features of studies of critical pathways
Study Setting Design,
Outcomes
Children with an asthma
exacerbation admitted to an
academic medical center: patients
randomized to pathway group or
usual care group 22
Patients undergoing hip or knee
replacement surgery at an academic
medical center in Australia: patients
randomized to pathway group or
usual care group 21
Patients being treated for hip fracture
at an academic medical center in
Australia 20
Patients presenting to the emergency
department with pneumonia at 19
medical centers across Canada:
hospitals were randomized to
pathway group or usual care group 23
Patients in need of supplemental
oxygen therapy: prospective beforeafter
analysis with washout period 26
Patients undergoing unilateral neck
dissection at an academic medical
center: pathway group compared to
historical controls and nonrandomized
concurrent controls 25
Patients admitted to a community
teaching hospital for the treatment of
an asthma exacerbation: pathway
group compared to historical controls
and non-randomized concurrent
controls 24
Patients undergoing one of several
defined surgical procedures over a
seven year period at an academic
medical center: pathway group
compared to historical controls as
well as to patients from other similar
hospitals in the same region 27
Level 1,
Level 1
Level 1,
Level 1
Level 1,
Level 2
Level 1,
Level 1
Level 2,
Level 1
Level 2,
Level 1
Level 2,
Level 1
Level 2,
Level 2
585
Clinical Outcomes
(p value)
Shorter duration of
intensive nebulizer
treatment (0.02), length
of stay (0.01)
Shorter length of stay
(0.011) and time to
mobilization (0.001) and
ambulation (0.02)
Shorter length of stay
(0.03)
Lower rates of bed-days
per patient (0.04),
shorter length of stay
(0.01) and shorter
duration of IV therapy
(0.01)
More frequent
appropriate oxygen
discontinuation orders
(0.001)
Length of stay decreased
compared to historical
controls (0.001) but not
concurrent controls
More appropriate
antibiotic use and
conversion to nebulizer
treatment (0.002 for
historical controls, 0.05
for concurrent controls)
Decreased lengths of
stay for the pathway
group compared to the
historical controls, but
not compared to the
concurrent control group
Non-Clinical
Outcomes (p value)
Lower room charges
(0.001) and therapy
charges (0.001)
Increased costs (0.02)
Total costs decreased
compared to historical
controls (0.001) but
not concurrent
controls

References
1. Every NR, Hochman J, Becker R, Kopecky S, Cannon CP. Critical pathways : a review.
Committee on Acute Cardiac Care, Council on Clinical Cardiology, American Heart
Association. Circulation. 2000; 101: 461-5.
2. Critical Path Software Smoothes Road for Automotive Supplier. Industrial Engineering.
1992; 24: 28-29.
3. Kallo G. The reliability of critical path method (CPM) techniques in the analysis and
evaluation of delay claims. Cost Engineering. 1996; 38: 35-37.
4. Hatfield MN, J. The case for critical path. Cost Engineering. 1998; 40: 17-18.
5. Pearson SD, Goulart-Fisher D, Lee TH. Critical pathways as a strategy for improving care:
problems and potential. Ann Intern Med. 1995; 123: 941-8.
6. Holmboe ES, Meehan TP, Radford MJ, Wang Y, Marciniak TA, Krumholz HM. Use of
critical pathways to improve the care of patients with acute myocardial infarction. Am J
Med. 1999; 107: 324-31.
7. Phillips KA, Shlipak MG, Coxson P, Goldman L, Heidenreich PA, Weinstein MC, et al.
Underuse of beta-blockers following myocardial infarction. JAMA. 2001; 285: 1013.
8. Bradshaw BG, Liu SS, Thirlby RC. Standardized perioperative care protocols and reduced
length of stay after colon surgery. J Am Coll Surg. 1998; 186: 501-6.
9. Dardik A, Williams GM, Minken SL, Perler BA. Impact of a critical pathway on the results
of carotid endarterectomy in a tertiary care university hospital: effect of methods on
outcome. J Vasc Surg. 1997; 26: 186-92.
10. Firilas AM, Higginbotham PH, Johnson DD, Jackson RJ, Wagner CW, Smith SD. A new
economic benchmark for surgical treatment of appendicitis. Am Surg. 1999; 65: 769-73.
11. Hanna E, Schultz S, Doctor D, Vural E, Stern S, Suen J. Development and implementation
of a clinical pathway for patients undergoing total laryngectomy: impact on cost and
quality of care. Arch Otolaryngol Head Neck Surg. 1999; 125: 1247-51.
12. Leibman BD, Dillioglugil O, Abbas F, Tanli S, Kattan MW, Scardino PT. Impact of a
clinical pathway for radical retropubic prostatectomy. Urology. 1998; 52: 94-9.
13. Mabrey JD, Toohey JS, Armstrong DA, Lavery L, Wammack LA. Clinical pathway
management of total knee arthroplasty. Clin Orthop. 1997; 113: 125-33.
14. Macario A, Horne M, Goodman S, Vitez T, Dexter F, Heinen R, et al. The effect of a
perioperative clinical pathway for knee replacement surgery on hospital costs. Anesth
Analg. 1998; 86: 978-84.
15. March LM, Cameron ID, Cumming RG, Chamberlain AC, Schwarz JM, Brnabic AJ, et al.
Mortality and morbidity after hip fracture: can evidence based clinical pathways make a
difference? J Rheumatol. 2000; 27: 2227-31.
16. Noedel NR, Osterloh JF, Brannan JA, Haselhorst MM, Ramage LJ, Lambrechts D. Critical
pathways as an effective tool to reduce cardiac transplantation hospitalization and charges.
J Transpl Coord. 1996; 6: 14-9.
17. Rumble SJ, Jernigan MH, Rudisill PT. Determining the effectiveness of critical pathways
for coronary artery bypass graft patients: retrospective comparison of readmission rates. J
Nurs Care Qual. 1996;11(2):34-40.
18. Warner BW, Kulick RM, Stoops MM, Mehta S, Stephan M, Kotagal UR. An evidencedbased
clinical pathway for acute appendicitis decreases hospital duration and cost. J
Pediatr Surg. 1998;33(9):1371-5.
586

19. Willis B, Kim LT, Anthony T, Bergen PC, Nwariaku F, Turnage RH. A clinical pathway
for inguinal hernia repair reduces hospital admissions. J Surg Res. 2000;88(1):13-7.
20. Choong PF, Langford AK, Dowsey MM, Santamaria NM. Clinical pathway for fractured
neck of femur: a prospective, controlled study Med J Aust. 2000;172(9):423-6.
21. Dowsey MM, Kilgour ML, Santamaria NM, Choong PF. Clinical pathways in hip and knee
arthroplasty: a prospective randomised controlled study. Med J Aust. 1999;170(2):59-62.
22. Johnson KB, Blaisdell CJ, Walker A, Eggleston P. Effectiveness of a clinical pathway for
inpatient asthma management. Pediatrics. 2000;106(5):1006-12.
23. Marrie TJ, Lau CY, Wheeler SL, Wong CJ, Vandervoort MK, Feagan BG. A controlled
trial of a critical pathway for treatment of community-acquired pneumonia. CAPITAL
Study Investigators. Community-Acquired Pneumonia Intervention Trial Assessing
Levofloxacin. JAMA. 2000;283(6):749-55.
24. Bailey R, Weingarten S, Lewis M, Mohsenifar Z. Impact of clinical pathways and practice
guidelines on the management of acute exacerbations of bronchial asthma. Chest.
1998;113(1):28-33.
25. Chen AY, Callender D, Mansyur C, Reyna KM, Limitone E, Goepfert H. The impact of
clinical pathways on the practice of head and neck oncologic surgery: the University of
Texas M. D. Anderson Cancer Center Experience. Arch Otolaryngol Head Neck Surg.
2000;126(3):322-6.
26. Wong C, Visram F, Cook D, Griffith L, Randall J, O'Brien B, et al. Development,
dissemination, implementation and evaluation of a clinical pathway for oxygen therapy.
CMAJ. 2000;162(1):29-33.
27. Pearson SD, Kleefield SF, Soukop JR, Cook EF, Lee TH. Critical pathways intervention to
reduce length of hospital stay. Am J Med. 2001;110:175-80.
587

588

Chapter 53. Clinical Decision Support Systems
Robert Trowbridge, MD
University of California, San Francisco School of Medicine
Scott Weingarten, MD, MPH
University of California, Los Angeles School of Medicine
Background
Integrating medical knowledge and advances into the clinical setting is often difficult due
to the complexity of the involved algorithms and protocols. Clinical decision support systems
(CDSS) assist the clinician in applying new information to patient care through the analysis of
patient-specific clinical variables. 1 Many of these systems are used to enhance diagnostic efforts
and include computer-based programs such as Dxplain that provide extensive differential
diagnoses based on clinical information entered by the clinician. 2 Other forms of clinical
decision support systems, including antibiotic management programs and anticoagulation dosing
calculators, seek to prevent medical errors and improve patient safety. 3-5
Practice Description
Clinical decision support systems vary greatly in their complexity, function and
application. 1 These clinical tools differ from practice guidelines (Chapter 51) and critical
pathways (Chapter 52) in that they require the input of patient-specific clinical variables and as a
result provide patient-specific recommendations. Guidelines and pathways, in contrast, may not
require the input of such information and provide more general suggestions for care and
treatment. Although many clinical decision support systems are now computer-based, some are
relatively simple, with no inherently complex internal logic systems. Among the most common
forms of support systems are drug-dosing calculators, computer-based programs that calculate
appropriate doses of medications after clinicians input key data (eg, patient weight, indication for
drug, serum creatinine). These calculators are especially useful in managing the administration
of medications with a narrow therapeutic index (see Chapter 9). More complex systems include
computerized diagnostic tools that, although labor intensive and requiring extensive patientspecific
data entry, may be useful as an adjunctive measure when a patient presents with a
confusing constellation of symptoms and an unclear diagnosis. Other systems, both simple and
complex, may be integrated into the point-of-care and provide accessible reminders to clinicians
regarding appropriate management based on previously entered data. These systems may be
most practical when coupled with computerized physician order entry and electronic medical
records (see Chapter 6). Finally, through their integration with practice guidelines and critical
pathways, decision support systems may provide clinicians with suggestions for appropriate
care, thus decreasing the likelihood of medical errors. For example, a guideline for the
management of community-acquired pneumonia may include a clinical tool that, after the input
of patient-specific data, would provide a recommendation regarding the appropriateness of
inpatient or outpatient therapy. 6
589

Prevalence and Severity of the Target Safety Problem/Opportunities for Impact
A great variety of patient safety issues may be affected by clinical decision support
systems. Since the systems are typically intercalated at the point-of-care, they are likely to have
their greatest effect on problems related to the individual patient. Some of the best studied
applications of decision support systems are those that seek to prevent adverse drug reactions
and inappropriate drug dosing as well as those that facilitate the appropriate use of effective
prophylactic measures, such as those for venous thromboembolic disease (see Chapter 31). 4,5
Decision support systems may also help ensure a minimum standard for quality of care as part of
the implementation of practice guidelines by providing patient-specific recommendations after
the input of certain clinical variables. This combination of decision support systems and
guidelines may be especially effective in conjunction with the use of the electronic medical
record. 7 To date no high quality studies have examined the impact of widespread institution of
decision support, but local studies suggest that it may be substantial. An evaluation of an
antibiotic management system in a single 12-bed Intensive Care Unit over a 3 year period
revealed that use of the system resulted in 24 fewer adverse drug reactions and 194 fewer cases
of antibiotic-susceptibility mismatch. 4 These results and others suggest that national
implementation of decision support systems could markedly improve patient safety.
Study Design
The functionality and effectiveness of clinical decision support systems has been
evaluated in a number of systematic reviews. The seminal review, last updated in 1998,
investigated the use of computer-based systems in all clinical settings. 3 The authors completed
an exhaustive search of electronic databases and the bibliographies of pertinent articles and
found 68 prospective trials on the subject. The vast majority of these studies (90%) were
randomized trials. The articles were rated on a 10-point scale for quality. The scale assessed the
design features of the trials, including method of randomization, baseline comparability of the
study groups, allocation unit, outcome measures, and degree of follow-up. The mean score on
this validity scale was 6.4 for studies published prior to 1992 and 7.7 for subsequent studies.
Another systematic review evaluated the use of computer-based decision aids in the
provision of outpatient preventive care. 8 This study, which was based on a search of electronic
databases including MEDLINE, found 16 trials that met the pre-defined inclusion criteria. Only
randomized controlled studies were included. Those that used only historical controls were
excluded. The acceptable studies were then evaluated using weighted mixed effects model
regression analysis.
Yet another systematic review investigated the utility of computer systems in the primary
care setting. 9 A detailed search of several electronic databases and a hand search of
bibliographies and conference proceedings yielded 30 studies that met criteria for inclusion.
These papers were ranked by the same validity score utilized in the systematic review of
computer-based decision supports systems described above. The average validity score of the
included trials was 6.7 on the 1-10 scale.
Most relevant to patient safety per se is a fourth systematic review of the utility of
computers in medication dosing. 10 This review, which located studies in the Cochrane
Collaboration on Effective Clinical Practice, was also based on an extensive search of electronic
databases, supplemented by a bibliography search and consultation with experts. Of 16 relevant
studies, all but one were randomized controlled clinical trials. Time series studies were not
included. The included trials were then evaluated using a random effects model.
590

A final systematic review analyzed the utility of computers in the implementation of
practice guidelines which, in such a setting, may be considered clinical decision support
systems. 11 A structured search of MEDLINE, CINAHL, bibliographies and books found 25
papers detailing the use of 20 such systems. In this group, there were 10 time series studies (all
without external controls) and 10 controlled trials, 9 of which were randomized.
Two well-designed studies have been published since the above systematic reviews. One
is a cluster randomized trial of the use of a decision support system in the treatment of patients
with hypertension. 12 The second is a prospective time series investigation of a decision support
system for the prevention of thromboembolic disease in post-surgical patients. 4
Study Outcomes
Most of the reviewed studies reported results in terms of patient outcomes and provider
performance. The most common clinical outcomes (Level 2) included the relative adherence to
specific recommendations and the degree to which prospectively described tenets of “appropriate
practice” were followed. Examples of such principles included whether clinicians followed
medication dosage recommendations and delivered appropriate preventive health care measures.
Several studies reported even more definitive clinical data (Level 1) including the degree of
blood pressure reduction, the incidence of adverse drug reactions, the control of postoperative
pain and percentage of patients with therapeutic drug blood levels.
Evidence for Effectiveness of Practice
The majority of the systematic reviews portray clinical decision support systems in a
positive light. The seminal systematic review of computer-based decision support systems, for
example, found that 43 of the 65 investigated studies showed at least some benefit in either the
process of care or patient outcomes. 3 Most impressively, 74% of the studies of preventive health
care reminder systems and 60% of the evaluations of drug dosing models reported a positive
impact. In addition, 6 of the 14 studies that reported actual patient outcomes reported a beneficial
effect. However only one of the five diagnostic aids, a model used to predict postoperative
pulmonary complications, showed encouraging results. The others, including systems designed
to aid diagnostic efforts in patients with chest and abdominal pain, were ineffective. This review
was rigorously conducted and only those papers with robust study design were included. Overall
it provides compelling evidence for the effectiveness of specific decision support systems.
These results are supported by the review of systems to aid the delivery of preventive
health care. 8 By applying weighted mixed effects model regression analysis to the 16 identified
studies, the authors found decision support systems resulted in significantly improved rates of
delivery of preventive health care (OR 1.77, 95% CI: 1.38-2.27). Computer systems, however,
were not statistically superior to manual reminder systems. Although this study reviewed many
of the same papers that were evaluated in the reviews discussed above, it provides further
evidence of the utility of these systems in improving the process of care.
The review of clinical decision support systems in the outpatient setting provided
similarly encouraging results. 9 All of the 21 studies that examined processes of care found
improvements. Results were somewhat less impressive in the articles reporting more definitive
clinical outcomes including level of blood pressure reduction and patient satisfaction. Only one
of three such studies found significant benefit.
The systematic review of the use of computer models to determine drug doses provides
some of the most compelling evidence for application of decision support systems to patient
safety and error avoidance efforts. 10 Seven of the 11 studies showed a beneficial effect on drug
591

dosing, and more significantly, 4 of 6 showed a decrease in adverse drug effects. Five of six also
reported direct positive effects on the patient, including one that reported improved control of
postoperative pain. Since the majority of the analyzed studies were randomized controlled trials,
this review provides powerful evidence that decision support systems may prevent medical
errors and other adverse events. A more recent study of an antibiotic management program also
supports this contention. 5 In this prospective before-after trial, the use of the decision support
system resulted in a substantial and significant decrease in the incidence of adverse drug
reactions (p=0.018) and prescriptions for medications contraindicated by allergy (p

Costs and Implementation
Very few studies specifically address the cost of developing and implementing decision
support systems. We posit that the costs may be substantial as the majority of systems are now
computer-based and require significant hardware, most commonly placed at the point-of-care. In
addition, many of the successful systems were integrated with computerized medical record
systems and physician order entry (Chapter 6), which are not yet in universal use. The
development and frequent updating of system software is also likely to be very expensive.
Despite these concerns, the widespread implementation of successful systems is feasible and will
likely become even more so as providers and systems increasingly shift to computerized medical
record systems.
Comment
The preponderance of evidence suggests that clinical decision support systems are at least
somewhat effective. Their highest utility has been demonstrated in the prevention of medical
errors, especially when coupled with a computerized medical record and directly intercalated
into the care process. Unlike more passive methods such as education and feedback, decision
support systems are generally able to modify physician behavior and affect the process of care.
Although the results of support systems have been far less positive when used in the ongoing
care of patients with chronic diseases or to help with diagnostic decision making, these
capabilities may improve with further technological advances.
References
1. Payne TH. Computer decision support systems. Chest. 2000;118:47S-52S.
2. Barnett GO, Cimino JJ, Hupp JA, Hoffer EP. DXplain. An evolving diagnostic decisionsupport
system. JAMA. 1987; 258: 67-74.
3. Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computer-based clinical decision
support systems on physician performance and patient outcomes: a systematic review.
JAMA. 1998; 280: 1339-46.
4. Durieux P, Nizard R, Ravaud P, Mounier N, Lepage E. A clinical decision support system
for prevention of venous thromboembolism: effect on physician behavior. JAMA. 2000;
283: 2816-21.
5. Evans RS, Pestotnik SL, Classen DC, Clemmer TP, Weaver LK, Orme JF, et al. A
computer-assisted management program for antibiotics and other antiinfective agents. N
Engl J Med. 1998; 338: 232-8.
6. Marrie TJ, Lau CY, Wheeler SL, Wong CJ, Vandervoort MK, Feagan BG. A controlled
trial of a critical pathway for treatment of community-acquired pneumonia. CAPITAL
Study Investigators. Community-Acquired Pneumonia Intervention Trial Assessing
Levofloxacin. JAMA. 2000; 283: 749-55.
7. Shiffman RN, Brandt CA, Liaw Y, Corb GJ. A design model for computer-based guideline
implementation based on information management services. J Am Med Inform Assoc.
1999; 6: 99-103.
8. Shea S, DuMouchel W, Bahamonde L. A meta-analysis of 16 randomized controlled trials
to evaluate computer-based clinical reminder systems for preventive care in the ambulatory
setting. J Am Med Inform Assoc. 1996; 3: 399-409.
9. Sullivan F, Mitchell E. Has general practitioner computing made a difference to patient
care? A systematic review of published reports. BMJ. 1995; 311: 848-52.
593

10. Walton R, Dovey S, Harvey E, Freemantle N. Computer support for determining drug
dose: systematic review and meta-analysis. BMJ. 1999; 318: 984-90.
11. Shiffman RN, Liaw Y, Brandt CA, Corb GJ. Computer-based guideline implementation
systems: a systematic review of functionality and effectiveness. J Am Med Inform Assoc.
1999; 6: 104-14.
12. Montgomery AA, Fahey T, Peters TJ, MacIntosh C, Sharp DJ. Evaluation of computer
based clinical decision support system and risk chart for management of hypertension in
primary care: randomised controlled trial. BMJ. 2000; 320: 686-90.
13. Montgomery AA, Fahey T. A systematic review of the use of computers in the
management of hypertension. J Epidemiol Community Health. 1998; 52: 520-5.
14. Weingarten MA, Bazel D, Shannon HS. Computerized protocol for preventive medicine: a
controlled self-audit in family practice. Fam Pract. 1989; 6: 120-4.
594

Chapter 54. Educational Techniques Used in Changing Provider Behavior
Robert Trowbridge, MD
University of California, San Francisco School of Medicine
Scott Weingarten, MD, MPH
University of California, Los Angeles School of Medicine
Background
A number of techniques have been used to modify the behavior of practicing physicians. 1
Continuing medical education, practice guidelines and critical pathways represent a major thrust
of these efforts. The relative effectiveness of each is largely dependent on the particular strategy
employed in their implementation. 2 Traditionally these strategies have focused on lectures and
printed materials but other techniques have also been utilized, including audit and feedback,
academic detailing, local opinion leaders and reminder systems. In addition, some have
championed the use of sentinel event reporting and root cause analysis in graduate medical
education programs. 3 Only recently have these various techniques been critically evaluated for
their effectiveness at changing physician behavior.
This chapter reviews the evidence regarding the utility of educational-oriented techniques
to improve provider behavior, particularly as they do or might pertain to patient safety practices.
Incident reporting (Chapter 4), root cause analysis (Chapter 5), guidelines (Chapter 51),
pathways (Chapter 52), and decision support systems (Chapter 53) are reviewed elsewhere in the
Report.
Practice Description
The passive dissemination of information through the use of lectures, conferences,
mailings and printed materials remains the primary method to alter physician behavior. This
primacy has not been substantially challenged in practice, although more interactive techniques
have increasingly been utilized. Academic detailing, for example, involves the process of having
invested and well-informed agents for change interacting with individual physicians to promote
certain tenets of practice. Alternatively, audit and feedback entails the review and return to the
clinician of their own process of care and patient outcomes (often compared with local or
national benchmarks or evidence-based standards) in the hopes it will result in more appropriate
medical care. 3 Reminder systems, which may be computerized and embedded in the electronic
medical record, prompt physicians to provide certain health care measures. They differ from
clinical decision support systems (Chapter 53) in that they may not provide information tailored
to the specific patient. Finally, opinion leaders, usually respected local physicians, may improve
4, 5
health care by championing “best practices” on a regional basis.
Prevalence and Severity of Target Safety Problem/Opportunities for Impact
It is well established that physicians are unable to keep abreast of the staggering volume
of published medical literature. This is reflected by the many studies that demonstrate the glacial
pace at which many beneficial advances are incorporated into medical practice. Practice
guidelines, clinical decision support systems and programs for physician education are potential
solutions to this problem, but their effectiveness is greatly dependent on the methods used in
their implementation. 2 Despite the presence of comprehensive guidelines on the treatment of
reactive airways disease, for example, a substantial percentage of asthmatic patients do not
595

eceive appropriate care. 6, 7 Physician education techniques that reliably impact practice patterns
may yield substantial improvements in patient care and safety.
Study Designs
The Cochrane Group completed a series of systematic reviews of physician education
based on the Research and Development Base in Continuing Medical Education, a central
database compiled from an extensive search of electronic databases and bibliographies and
supplemented by contact with experts in the field. Although the initial review was completed in
1997, reviews are regularly updated as more pertinent data are published. One such study
evaluated the role of audit and feedback and found 37 randomized controlled studies comparing
this technique to non-interventional control groups. 3 An ancillary study by the Cochrane Group
compared audit and feedback with other educational strategies and located 12 randomized
controlled studies for analysis. 10 A second study of the effectiveness of audit and feedback was
completed by a separate group that searched MEDLINE and selected bibliographies for trials
investigating the strategy’s utility in improving immunization rates. Fifteen studies were
identified for inclusion, 5 of which were randomized controlled studies with 6 interrupted time
series evaluations and 4 before-after trials. 11 A third meta-analysis of peer-comparison feedback
systems used an extensive electronic database and bibliography search to locate 12 randomized
controlled studies. 12
The Cochrane Group also investigated the utility of academic detailing and found 18
randomized controlled studies. 13 A similar evaluation of local opinion leaders yielded 8
randomized controlled trials. 5
A separate Cochrane review was completed on the utility of printed educational materials
using the Cochrane Effective Practice and Organization of Care Group database. The search of
this database, which was compiled in the same manner as the Research and Development Base in
Continuing Medical Education, found 10 randomized controlled trials and one interrupted time
series study fulfilling criteria for analysis. 14
Study Outcomes
Few of the studies report outcomes specific to the field of patient safety. The vast
majority are concerned with process of care rather than the outcomes of care. Although clinical
outcomes are reported in at least one of the studies evaluated in each of the systematic reviews
(with one exception), the majority relate outcomes pertaining to physician performance. Some of
the more commonly described variables include the rates of appropriate provision of preventive
care measures and of adherence to appropriate treatment or diagnostic protocols.
Evidence for Effectiveness of Practice
Much of the evidence for the effectiveness of educational and implementation techniques
is of fair quality and the results are generally consistent across the various systematic reviews.
However, methodologic concerns prevented the completion of quantitative data synthesis in the
majority of the reviews. The studies are summarized in Table 54.1.
The initial comprehensive review found overall beneficial effect for 62% of
interventions. In investigations of effect on patient outcomes, 48% had favorable results.
Academic detailing and the use of local opinion leaders were the most effective techniques
evaluated. Physician reminder systems were also effective, as 22 of the 26 evaluated studies
revealed some benefit. The technique of audit and feedback was of marginal effectiveness and
conferences and printed materials were found to be relatively ineffective. Of note, multifaceted
596

interventions with at least 3 components were associated with a 71% success rate. 8 The second
comprehensive review of 102 randomized controlled studies supported these conclusions. Yet it
emphasized that the degree of effect with even the most consistently effective techniques was
moderate at best, and that the process of care rather than the outcome of care was the most
readily influenced variable. 9
The Cochrane reviews reported similar results. Audit and feedback was found to be
effective in 62% of the studies in which it was compared with non-interventional controls, but
the effect was typically small. The results were not substantially different when audit and
feedback was augmented by conferences or educational materials or was part of a multifaceted
intervention. 3 In the review of comparative trials, however, this technique was found to be
inferior to reminder systems in 2 of the 3 trials where a direct comparison was made. 10 The
second review of audit and feedback, which focused on improving immunization rates, found
beneficial results in 4 of 5 randomized controlled trials evaluated. Statistically significant
changes were present in at least 2 of these evaluations. However, the marginal effect was small
and likely was overwhelmed by the relatively high cost of the intervention. 11 Finally the metaanalysis
of the 12 randomized controlled trials investigating peer-comparison feedback systems
did establish a modest benefit for the use of audit and feedback (p

Costs and Implementation
Although the cost-effectiveness of the various educational techniques has not been
explicitly studied, it is clear that several may require substantial outlay in terms of financial
resources and personnel. It also appears that the forms of education that are most effective,
including academic detailing and local opinion leaders, are also the most expensive to design and
support. Programs of printed materials and lectures, although dramatically less effective, are
substantially less expensive to implement. It is unclear whether the integration of Internet
technology and computer-based education initiatives will result in substantial changes in efficacy
or cost. Finally, the relative cost-effectiveness of the various techniques remains unclear.
Comment
From studies of randomized controlled trials, it appears that academic detailing and local
opinion leaders are frequently associated with at least some benefit. Reminder systems are also
effective in specific situations and the utility of audit and feedback has been established,
although unimpressively. Traditional programs of conferences, lectures and printed materials are
ineffective at inducing changes in physician behavior. None of the current techniques, however,
have demonstrated a consistent ability to induce substantial and durable changes in physician
behavior. The relative cost-effectiveness of the various techniques is uncertain; it remains
unclear if the added cost of the more effective strategies (ie, academic detailing and local opinion
leaders) is justified given their relatively small marginal increase in effectiveness. Finally there
are few data regarding the specific utility of these techniques in increasing patient safety and/or
the prevention of medical errors. However, techniques effective in other areas of medicine are
likely to be equally effective in inducing practices changes to improve patient safety.
598

Table 54.1. Studies of techniques for changing physician behavior*
Study Setting Study
Design
Review of 99 trials assessing the
effect of educational techniques on
physician performance in all clinical
settings 8
Review of 102 trials assessing the
effect of educational techniques on
physician performance in all clinical
settings 9
Review of 37 randomized controlled
trials of the utility of audit and
feedback in all clinical settings in the
US, Europe and Australia 3
Review of 12 trials comparing the
effect of audit and feedback with
other educational techniques on 2194
physicians in all clinical settings 10
Review of fifteen studies, of which
five were randomized controlled
trials, investigating the use of audit
and feedback in improving
immunization rates in adults and
children in the outpatient setting in
the US and the UK 11
Review of 18 trials investigating the
effects of academic detailing on 1896
physicians in the US, Canada,
Europe, Indonesia and Australia 13
Review of 8 randomized controlled
trials investigating the effect of local
opinion leaders on 296 physicians in
the US, Canada and Hong Kong 5
Review of eleven studies evaluating
the effect of printed education
materials on over 1848 physicians in
a variety of clinical settings 14
599
Results
Level 1A 62% of the interventions were associated with
beneficial results; academic detailing, local opinion
leaders and reminder systems were the most
effective while audit and feedback was less so;
traditional CME programs were ineffective
Level 1A Academic detailing and local opinion leaders were
the most effective techniques; audit and feedback
and reminder systems were less effective;
multifaceted approaches were effective, especially
at influencing the process of care
Level 1A Eight of 13 studies showed a moderate beneficial
effect with audit and feedback with little change
noted when other interventions were added or a
multifaceted approach was used
Level 1A Two of 3 trials showed reminder systems
outperformed audit and feedback; 4 studies
demonstrated little benefit to adding other
modalities to audit and feedback
Level 1A Twelve of the 15 studies showed a benefit with
audit and feedback and of the 5 RCTs, 4 showed a
beneficial trend that was significant in at least 2 of
the trials
Level 1A All of the evaluated studies showed some degree of
benefit although only one looked specifically at
patient outcomes
Level 1A Six of the 7 studies evaluating effects on physician
performance showed a beneficial effect with 2 of
these being statistically significant; one of the 3
trials evaluating patient outcomes showed a
significantly positive effect
Level 1A None of the studies reported significantly improved
outcomes with the use of printed educational
materials
* CME indicates continuing medical education; RCT, randomized controlled trial.

References
1. Smith WR. Evidence for the effectiveness of techniques To change physician behavior.
Chest. 2000; 118: 8S-17S.
2. Grimshaw JM, Russell IT. Effect of clinical guidelines on medical practice: a systematic
review of rigorous evaluations. Lancet. 1993; 342: 1317-22.
3. Thomson O'Brien MA, Oxman AD, Davis DA, Haynes RB, Freemantle N, Harvey EL.
Audit and feedback: effects on professional practice and health care outcomes. In: The
Cochrane Library, Issue 2, 2000:Oxford: Update Software.
4. Szilagyi PG, Bordley C, Vann JC, Chelminski A, Kraus RM, Margolis PA, et al. Effect of
patient reminder/recall interventions on immunization rates: A review. JAMA. 2000; 284:
1820-7.
5. Thomson O'Brien MA, Oxman AD, Haynes RB, Davis DA, Freemantle N, Harvey EL.
Local opinion leaders: effects on professional practice and health care outcomes. In: The
Cochrane Library, Issue 2, 2000. Oxford: Update Software..
6. Apter AJ, Van Hoof TJ, Sherwin TE, Casey BA, Petrillo MK, Meehan TP. Assessing the
quality of asthma care provided to Medicaid patients enrolled in managed care
organizations in Connecticut. Ann Allergy Asthma Immunol. 2001; 86: 211-8.
7. Rabe KF, Vermeire PA, Soriano JB, Maier WC. Clinical management of asthma in 1999:
the Asthma Insights and Reality in Europe (AIRE) study. Eur Respir J. 2000; 16: 802-7.
8. Davis DA, Thomson MA, Oxman AD, Haynes RB. Changing physician performance. A
systematic review of the effect of continuing medical education strategies. JAMA. 1995;
274: 700-5.
9. Oxman AD, Thomson MA, Davis DA, Haynes RB. No magic bullets: a systematic review
of 102 trials of interventions to improve professional practice. CMAJ. 1995; 153: 1423-31.
10. Thomson O'Brien MA, Oxman AD, Davis DA, Haynes RB, Freemantle N, Harvey EL.
Audit and feedback versus alternative strategies: effects on professional practice and health
care outcomes. In: The Cochrane Library, Issue 2, 2000. Oxford: Update Software.
11. Bordley WC, Chelminski A, Margolis PA, Kraus R, Szilagyi PG, Vann JJ. The effect of
audit and feedback on immunization delivery: a systematic review. Am J Prev Med. 2000;
18: 343-50.
12. Balas EA, Boren SA, Brown GD, Ewigman BG, Mitchell JA, Perkoff GT. Effect of
physician profiling on utilization. Meta-analysis of randomized clinical trials. J Gen Intern
Med. 1996; 11: 584-90.
13. Thomson O'Brien MA, Oxman AD, Davis DA, Haynes RB, Freemantle N, Harvey EL.
Educational outreach visits: effects on professional practice and health care outcomes. In:
The Cochrane Library, Issue 2, 2000. Oxford: Update Software.
14. Freemantle N, Harvey EL, Wolf F, Grimshaw JM, Grilli R, Bero LA. Printed educational
materials: effects on professional practice and health care outcomes. In: The Cochrane
Library, Issue 2, 2000. Oxford: Update Software.
600

Chapter 55. Legislation, Accreditation, and Market-Driven and Other
Approaches to Improving Patient Safety
Robert Trowbridge, MD
Robert M.Wachter, MD
University of California, San Francisco School of Medicine
Background
Although conventional approaches to health care quality improvement based on
educating providers and offering feedback have been proposed, the tremendous public and
professional concerns raised by the Institute of Medicine’s report on medical errors 1 have led to
an unusual amount of interest in regulatory and legislative efforts to improve safety. Given the
traditional American dependence on education, competition, and other non-coercive mechanisms
of change, the shift toward a regulatory approach is evidence of the depth of concern this issue
has engendered. As these regulatory and legislative approaches are considered, it is also
worthwhile to consider the role of health care payers, hospital accreditation organizations, and
professional societies, all of whom have also led safety initiatives. 2-4 This chapter considers the
potential advantages and disadvantages of legislative, regulatory, professional society, and
market-oriented approaches to implementing patient safety efforts, and reviews the evidence
regarding their effectiveness.
Government Legislation and Regulation
The concept of patient safety has been championed by several prominent legislators in
both major political parties and has become the topic of a great deal of national debate. Proposals
to date include the establishment of voluntary and mandatory error reporting systems, the
publication of outcomes data, and the development of several Health Care Financing
Administration (HCFA) programs to prevent medical errors. 5 In addition, in 2000 the Agency for
Healthcare Research and Quality (AHRQ) established a Center for Quality Improvement and
Patient Safety, whose mandate in part is to fund research and demonstration projects in patient
safety. Though most of these Federal efforts are in the formative stages, 6 successful Federal
agency and regulatory efforts outside of medicine, most notably in workplace safety and
commercial airline travel, may herald medicine’s success. In these fields, the Occupational
Safety and Health Administration (OSHA) and the Federal Aviation Administration (FAA) are
7, 8
credited with catalyzing significant improvements in safety over the past 30 years.
Despite a paucity of data regarding the effect of State legislation on medical errors and
patient safety, there is some evidence regarding the effectiveness of State regulatory efforts to
improve health care quality. 9-11 Among the most prominent set of regulations are the New York
State enactments as a result of the death of Libby Zion.
The daughter of a prominent reporter, Ms. Zion died soon after being admitted to the
medical service of a New York City hospital. A detailed investigation of the circumstances of
her death subsequently raised concerns regarding working conditions and the supervision of
resident physicians. The Bell Commission, formed at the behest of the New York State Health
Commissioner, later recommended major reform in these areas and several regulations were
enacted mandating dramatic changes in working conditions and supervisory oversight. 12
These regulations markedly altered resident physician education in New York State.
Although they anecdotally resulted in an improvement in resident morale and quality of life, 13
601

their effect on patient safety is less certain. One retrospective cohort study demonstrated that
patients treated after the work-hour limitations were instituted were more likely to suffer from
complications and delays in the performance of diagnostic tests. 11 A second retrospective
analysis of patient transfer-of-care, a bi-product of restricting resident work-hours, showed
increased lengths of stay and utilization of laboratory testing for patients that were “handed
off.” 10 A third study, however, revealed exactly contradictory results. It found the work-hours
limitations led to shorter lengths of stay and fewer medication errors. 9 These studies have been
criticized for concentrating on the work-hour regulations when the main finding of the Bell
Commission was that increased supervision of resident physicians was a more important
initiative. 14 (See Chapter 46 for a more complete discussion of fatigue and work hours).
Although regulations to improve patient safety might be a more efficient way of
changing practice than less coercive methods such as education or feedback, the use of
government regulation in the assurance of patient safety has limitations. A primary concern is
that regulations may be crafted by legislators who lack intimate knowledge of the health care
system. In addition, health care in the United States is extremely heterogeneous - what may be
feasible and appropriate in one setting may be inapplicable in another. The differences between
the delivery of care in urban and rural settings may be particularly troublesome in this regard.
Finally, it is unclear whether government agencies would provide adequate funding to assist
health care organizations to comply with new regulations. For example, a cash-starved institution
faced with a resident work-hours mandate might need to decrease nurse staffing or defer
purchase of a computerized order entry system to meet the mandate.
Government agencies may also influence patient safety practices through means other
than direct legislation. For example, the National Nosocomial Infections Surveillance (NNIS)
system may serve as a template for the development of a government-sponsored safety program.
Established and administered by the Centers for Disease Control and Prevention (CDC), the
NNIS is a voluntary nationwide consortium of 300 hospitals that regularly report the incidence
of specified nosocomial infections. 15 Through analysis of the aggregate data, benchmarks are set
for the expected rates of nosocomial infection that hospitals may then strive to meet or better.
There is some evidence that the NNIS has contributed to a substantial decline in the rate of
nosocomial infections over the past several decades. 16 It is conceivable that a similar program
could be established for broader patient safety issues. Although voluntary and lacking
enforcement power, NNIS-like patient safety benchmarking could significantly improve safety,
especially if the data were made available to the public or accreditation agencies.
Infection control officers, another aspect of the CDC-championed infection control
movement, may also be applicable to the patient safety movement. Presently, infection control
officers employed by medical centers focus on improving the system and changing practices to
decrease the institutional rates of serious infections. An institutional “patient safety officer”
might assume an analogous function with regard to decreasing adverse events and errors. The
establishment of a patient safety officer or committee is one of the new Joint Commission on
Accreditation of Healthcare Organization’s (JCAHO) safety standards (discussed below). To
date, there are no data regarding the effectiveness of these practices in patient safety.
Accreditation Organizations
Accreditation organizations represent another sector of the health care industry that is
assuming new responsibilities in the field of patient safety. JCAHO, an outgrowth of
accreditation programs initiated by the American College of Surgeons several decades ago, is the
best known of these organizations. JCAHO conducts meticulous inspections of medical centers,
602

hospitals and other health care institutions on a triennial basis. Survey results are subsequently
used in the process of obtaining Medicare certification and State licensure and to obtain bond
ratings and managed care contracts. Although such inspections had previously included some
elements relating to the field of patient safety (including infection control and the prevention of
medication errors), they tended to focus on organizational topics including information
management, institutional leadership and strategic planning. 17 In response to concerns regarding
patient safety, however, JCAHO has recently launched a major patient safety initiative and
implemented an entirely new set of standards in July 2001. 18
The new JCAHO standards place a much greater emphasis on the prevention of medical
errors and the process of responding to medical errors once they occur. A particular focus of the
new initiative is the development of organization-specific patient safety programs. Such
programs, which will undoubtedly require substantial resources to implement, are expected to
have well-defined leadership, to proactively determine areas where errors are likely to occur, and
to be capable of effecting significant organizational change when necessary. The key elements of
these standards are listed in Table 55.1.
Although there is no published evidence that the patient safety standards previously
required by JCAHO have reduced medical errors, it seems reasonable to assume they have had
some salutary effect. 6 Because JCAHO reports are now publicly available and are used by a wide
variety of credentialing agencies and health care purchaser organizations, many hospitals and
other medical institutions will make serious and concerted efforts to meet the new standards.
Although lacking the enforcement power of Federal regulations, the agencies’ knowledge of, and
contacts within the medical community may produce change more efficiently. Moreover, the
process of accreditation involves frequent site visits between the agencies and health care
organizations. These visits, in turn, may allow for interactions between institutions and
accreditors which, under ideal circumstances, could allow for user input in to modifications of
regulations. How often this ideal is realized is not known, and some observers have questioned
JCAHO’s overall effectiveness. 19-21
Health Care Purchaser Initiatives
The business community has also reacted to the perceived crisis of safety in health care.
The most prominent example is the Leapfrog Group. 3 Sponsored by a consortium of major
corporate chief executive officers known as the Business Roundtable, the Leapfrog Group’s
stated commitment is “to mobilize employer purchasing power to initiate breakthrough
improvements in the safety of health care for Americans.” 22 Large volume purchasers of health
care, including Aetna, ATT, IBM and many other Fortune 500 companies, have joined the group.
Combined, their annual health care outlay is over $45 billion. Using their considerable financial
influence, the group hopes to impact medical care by requiring or incentivizing health care
providers to adhere to certain standards for the process and delivery of care. 22
Thus far the Leapfrog consortium has chosen to promote 3 patient safety practices: the
use of computerized physician order entry (Chapter 6), the involvement of critical care
physicians in the care of intensive care unit patients (Chapter 38), and the use of evidence-based
hospital referral systems (Chapter 18). 23 The latter practice refers to the referral of patients to
hospitals with the highest volume and best outcome figures for certain elective medical
procedures and treatments. These initiatives were selected because there is substantial evidence
that they enhance quality of care and their implementation is both feasible and easily assessed.
Initial research sponsored by the group suggests that implementing just these 3 strategies could
prevent almost 60,000 deaths per year and avoid over 500,000 errors in medication
603

administration. 24, 25 It is anticipated that other quality-related practices (some directed at patient
safety targets) will be added to the list as evidence for their effectiveness is accumulated.
The Leapfrog Group has also begun to outline a program to improve compliance with the
target practices. Plans include rewarding complying providers with an increased number of
patients, increasing remuneration for specific services, and providing public recognition. General
quality will be assessed and publicized through the use of rankings, including those assigned by
JCAHO and other accreditation organizations. 23 In addition, the employees of participating
companies will be encouraged to become more active in choosing providers that meet the
Leapfrog standards.
Although the Leapfrog initiative is in its nascent stages and there is presently no
objective evidence that it will favorably impact patient safety, the sheer financial clout of the
involved corporations may catalyze rapid and meaningful change. As with government
regulation, it is unclear which sector of the health care industry will bear the brunt of the
implementation costs. The institution of computerized order entry systems, for example, will
require substantial financial outlays (several million dollars in hardware, software and training
for the average hospital; see Chapter 6) that hospitals and medical groups may find extremely
difficult to absorb without assistance. Additionally, physicians and other health care providers
may resist changes forced upon them from outside medicine, or may “game the system” to create
the appearance of change simply to meet the standards. 26 Finally, purchaser initiatives may
create disparities in the level of patient safety among socioeconomic groups if the changes they
promote are only required of health care institutions that provide care to insured members of the
group.
Other Approaches
Professional Societies
Some of the earliest efforts to improve patient safety were actually directed by medical
professional societies rather than outside regulators, accreditors, or legislative bodies. The
American Society of Anesthesiologists (ASA), for example, formed the Anesthesia Patient
Safety Foundation in 1984 and has since promulgated a number of reforms that have
substantially changed the routine practice of anesthesia. Safety measures such as continuous
electrocardiographic monitoring, pulse oximetry and preoperative evaluation were strongly
championed by these organizations through the dissemination of quality standards and
newsletters. 1 While no evidence directly links these initiatives to improved patient safety, there
is little doubt that these reforms resulted in substantial advances. In fact, the standards of care
promoted by the ASA have been widely adopted and now represent the recognized minimum
level of appropriate care. 4 It is nonetheless difficult to separate the effects of these standards
from secular trends associated with technological and clinical advances.
Although medical professional societies do not possess the regulatory might of the
Federal government or the financial power of large health care purchasers, their standing in the
medical community and collective clinical experience are great advantages. It is well established
that physicians are more apt to follow practice guidelines sponsored by respected medical
societies than those issued by the government or industry. 27 Society-based programs are also
likely to allow for more provider input and may be more amenable to frequent modification.
Unfortunately, because they lack the power to compel, society recommendations cannot be the
sole agent of change, especially when the evidence supporting a practice is extremely strong and
the stakes are high. It is important to recognize that society recommendations carry a potential
604

for bias, particularly when the recommended changes may have an economic impact on society
members.
Publication of Performance Data
Each of the large entities described above (legislatures, accrediting bodies, purchasers,
payers, and professional societies) may choose to disseminate performance data to the public as
part of its quality improvement strategy. To date, most such report cards have focused on
discrete quality outcomes for single diseases or procedures (eg, mortality after coronary
angioplasty) rather than patient safety targets such as error or nosocomial infection rates.
Nonetheless, implicit in the vigorous debate regarding mandatory error reporting systems is the
question of whether public reporting of performance data is effective in either motivating
provider change or facilitating informed choices by patients.
Marshall and colleagues recently reviewed the evidence regarding the impact of public
reporting systems. 28 They found that such systems had a relatively small impact on patients (the
potential users of the data), but a greater impact on the hospitals (the sources of the data). They
posit that the impact is growing as the public becomes increasingly comfortable with both the
concept and interpretation of quality reports. Although some have claimed that public reporting
systems, such as New York State’s Cardiac Surgery Data System (which has reported riskadjusted
coronary bypass outcomes since 1990), have led to major improvements in quality, 29
this remains controversial. 30 Proponents point to New York’s falling bypass mortality rate as
evidence of the value of public reporting, but there is some evidence that the fall in the rate was
due to outmigration of high-risk patients to other states rather than true quality improvement. 31
Comment
Concerns about medical errors have spurred many organizations and institutions to
launch major patient safety initiatives. Perhaps because they represent a relatively recent
development or because empirical measurement of such macro-changes is difficult (isolating the
effects of individual interventions vs. other confounding influences), there is little objective
evidence either to determine if they will result in meaningful change or to consider their relative
advantages and disadvantages (Table 55.2). Yet Federal and State governments, accreditation
agencies such as JCAHO, and health care purchasers such as the Leapfrog Coalition, in
combination and independently, may eventually be highly effective champions of patient safety
initiatives. In addition, professional medical societies, given their influential role in the medical
community, are effective agents of change in certain circumstances. The work and involvement
of these diverse, powerful organizations and institutions may prove to be valuable adjuncts to the
more traditional mechanisms of change represented by practice guidelines, continuing medical
education programs, and decision support systems.
605

Table 55.1. New JCAHO safety standards
-Development of a leadership individual or group to devise and implement a
comprehensive patient safety program
-Development of a proactive error prevention program that includes means of identifying
potentially high risk areas
-Development of systems for the reporting of errors
-Development of an error-response system including protocols for root cause analysis
-Requirement for an annual report discussing errors, the response to errors and the
programs initiated to prevent future errors
-Requirement for hospital leaders to set “measurable objectives” for patient safety
programs
-Requirement for educational initiatives for employees, stressing the concept of patient
safety
606

Table 55.2. A comparison of non-local methods to promote patient safety practices
Approach Example Advantages Disadvantages
Legislation “Libby Zion”
laws limiting
resident work
hours
Accreditation JCAHO patient
safety standards
Market-based The Leapfrog
Group
Professional
Societies
Anesthesia
Patient Safety
Foundation
-potential for widespread
implementation
-supported by government
enforcement ability
-may be more flexible and
more easily modified than
legislation
-implemented at the level of
the health care organization
-health care providers may
have the opportunity for input
-uses the power of the market
to induce change (may be
more acceptable for many
providers than regulatory
solutions)
-may involve carrot (eg, higher
payments for better practices
or outcomes) rather than stick
alone to achieve impact
-readily accepted by health
care providers
-developed by providers
themselves, leading to better
“buy in”
-more easily modified when
new evidence or changes in
practice emerge
607
-inflexible
-limited acceptance by health
care providers
-potential to be developed with
inadequate input from providers
and experts
-may be politically driven with
limited applicability
-may not provide for costs of
implementation, leading to costshifting
away from other
beneficial patient safety
practices
-dependent on voluntary
participation in the accreditation
process
-limited enforcement ability
-generally assessed only every
few years
-potential to cause disparity in
care among groups not covered
by initiatives
-limited acceptance by health
care providers
-potential for standards to
develop with inadequate input
from health care providers
-change is not required, and
therefore implementation may
be limited
-minimal enforcement potential;
depends largely on voluntary
participation by practitioners
-potential for bias by
professional societies

References
1. Kohn LT, Corrigan JM, Donaldson MS. To Err is Human: Building a Safer Health System.
Washington, DC; Committee on Quality of Health Care in America: Institute of Medicine;
2000
2. Jencks SF. The government's role in hospital accountability for quality of care. Jt Comm J
Qual Improv. 1994; 20: 364-9.
3. Milstein A, Galvin RS, Delbanco SF, Salber P, Buck CR. Improving the safety of health
care: the leapfrog initiative. Eff Clin Pract. 2000; 3: 313-6.
4. Gaba DM. Anaesthesiology as a model for patient safety in health care. BMJ. 2000; 320:
785-8.
5. Parsons DW. Federal legislation efforts to improve patient safety. Eff Clin Pract. 2000; 3:
309-12.
6. Brennan TA. The role of regulation in quality improvement. Milbank Q. 1998; 76: 709-31,
512.
7. Occupational Safety and Health Administration; About OSHA. Available at:
http://www.osha.gov/as/opa/osha30yearmilestones.html. Accessed May 8, 2001.
8. NTSB Aviation Accident/Incident Database: Accidents and Accident Rates by NTSB
Classification Federal Aviation Administration. Available at:
http://nasdac.faa.gov/asp/fw_ntsb.asp. Accessed May 8, 2001.
9. Gottlieb DJ, Parenti CM, Peterson CA, Lofgren RP. Effect of a change in house staff work
schedule on resource utilization and patient care. Arch Intern Med. 1991; 151: 2065-70.
10. Lofgren RP, Gottlieb D, Williams RA, Rich EC. Post-call transfer of resident
responsibility: its effect on patient care. J Gen Intern Med. 1990; 5: 501-5.
11. Laine C, Goldman L, Soukup JR, Hayes JG. The impact of a regulation restricting medical
house staff working hours on the quality of patient care. JAMA. 1993; 269: 374-8.
12. DeBuono BA, Osten WM. The medical resident workload: the case of New York State.
JAMA. 1998; 280: 1882-3.
13. Piver CJ. Residency training under New York's new regulations. N Engl J Med. 1991; 324:
273-4.
14. Bell BM. Supervision, not regulation of hours, is the key to improving the quality of
patient care. JAMA. 1993; 269: 403-4.
15. National Nosocomial Infections Surveillance (NNIS) System report, data summary from
January 1990-May 1999, issued June 1999. Am J Infect Control. 1999; 27: 520-32.
16. Monitoring hospital-acquired infections to promote patient safety–United States, 1990-
1999. MMWR Morb Mortal Wkly Rep. 2000; 49: 149-53.
17. Braun BI, Koss RG, Loeb JM. Integrating performance measure data into the Joint
Commission accreditation process. Eval Health Prof. 1999; 22: 283-97.
18. Joint Commission on Accreditation of Healthcare Organizations, Standards, Standards
Revisions for 2001, Leadership Chapter. Available at:
http://www.jcaho.org/standards_frm.html. Accessed May 3, 2001.
19. Weiner CL. The Elusive Quest: Accountability in Hospitals. New York: Aldine De
Gruyter; 2000.
20. Dame L, Wolfe SM. The failure of "private" hospital regulation: An analysis of the Joint
Commission on Accreditation of Healthcare Organizations' inadequate oversight of
hospitals. Washington DC: Public Citizen Health Research Group; 1996.
608

21. Yessian M. The external review of hospital quality: the role of accreditation (OEI-01-97-
00051). Boston: Office of Evaluations and Inspections, Office of Inspector General,
Department of Health and Human Services; 1999.
22. The Leapfrog Group for Patient Safety: About Us. Available at:
http://www.leapfroggroup.org/about.htm. Accessed May 3, 2001.
23. The Leapfrog Group. The Leapfrog Group for Patient Safety: Patient Safety. Available at:
http://www.leapfroggroup.org/safety1.htm. Accessed May 3, 2001.
24. Young MP, Birkmeyer JD. Potential reduction in mortality rates using an intensivist model
to manage intensive care units. Eff Clin Pract. 2000; 3: 284-9.
25. Birkmeyer JD B, CM, Wennberg DE, Young M. The Leapfrog Group. Leapfrog Patient
Safety Standards: The Potential Benefit of Univeral Adoption. Available at:
http://www.leapfroggroup.org/PressEvent/Birkmeyer_ExecSum.PDF. Accessed May 3,
2001.
26. Simborg DW. DRG creep: a new hospital-acquired disease. N Engl J Med. 1981; 304:
1602-4.
27. Hayward RS, Guyatt GH, Moore KA, McKibbon KA, Carter AO. Canadian physicians'
attitudes about and preferences regarding clinical practice guidelines. CMAJ. 1997; 156:
1715-23.
28. Marshall MN, Shekelle PG, Leatherman S, Brook RH. The public release of performance
data: what do we expect to gain? A review of the evidence. JAMA. 2000; 283: 1866-74.
29. Mukamel DB, Mushlin AI. Quality of care information makes a difference: an analysis of
market share and price changes after publication of the New York State Cardiac Surgery
Mortality Reports. Med Care. 1998; 36: 945-54.
30. Hannan TJ. Detecting adverse drug reactions to improve patient outcomes. Int J Med Inf.
1999; 55: 61-4.
31. Omoigui NA, Miller DP, Brown KJ, Annan K, Cosgrove D, 3rd, Lytle B, et al.
Outmigration for coronary bypass surgery in an era of public dissemination of clinical
outcomes. Circulation. 1996; 93: 27-33.
609

610

PART V. ANALYZING THE PRACTICES
Chapter 56. Methodology for Summarizing the Evidence for the Practices
Chapter 57. Practices Rated by Strength of Evidence
Chapter 58. Practices Rated by Research Priority
Chapter 59. Listing of All Practices, Categorical Ratings, and Comments
611

612

Chapter 56. Methodology for Summarizing the Evidence for the Practices
Background
The Agency for Healthcare Research and Quality (AHRQ) charged the UCSF-Stanford
Evidence-based Practice Center with the task of rating or grading the patient safety practices
identified and evaluated in this Report. The Report is an anthology of diverse and extensive
patient safety practices, grouped by general topic (Part III, Sections A-H), and then further sorted
within chapters by individual practices. Synthesis was challenging, but critical in order that
readers and health care decision-makers could make judgments about practices to implement
and/or research further. Keeping these two audiences in mind, we set 3 goals for “rating” the
practices, as follows:
• Develop a framework for rating the main elements of practices within the
constraints of the available literature, providing as much pragmatic
information as possible to decision-makers who might endorse practices or
fund further research in the area of patient safety interventions;
• Document the limitations of the rating method so that those “taking home”
messages from the report understand the inherent limitations of making
comparisons of a highly heterogeneous field of possible practices; and
• Convey the results of the ratings in an organized, visually appealing,
accessible way that ensures that our cautionary notes regarding
oversimplifying the ratings are clear.
Ultimately, we aimed to weight the practices, based on the evidence, on a range of
dimensions, without implying any ability to calibrate a finely gradated scale for those practices
in between. Proper metrics for these comparisons (eg, cost-effectiveness analysis) require more
data than are currently available in the literature.
Data Inputs into the Practice Ratings
For each practice, information about various inputs into the final “roll-up”, as we referred
to our scoring of the practices, were prospectively determined. The decision about what
information to attempt to gather was based on the potential expected uses of summary tables of
practices. Three major categories of information were gathered to inform the rating exercise:
• Potential Impact of the Practice: based on prevalence and severity of the
patient safety target, and current utilization of the practice
• Strength of the Evidence Supporting the Practice: including an assessment of
the relative weight of the evidence, effect size, and need for vigilance to
reduce any potential negative collateral effects of practice implementation
• Implementation: considering costs, logistical barriers, and policy issues
Further clarification of the 3 categories is in order. Authors were asked to report on
prevalence and severity of the safety target for a given practice in order to categorize the
potential impact of implementing the practice. We added to this an assessment of the practice’s
potential impact by reviewing evidence of its current utilization. If an intervention is already
widely used, the room for improvement, stated in terms of additional reductions in adverse
events targeted by the practice that could be achieved by wider implementation, is less than if
few are currently using the practice. Thus, potential impact of implementing the practice is a
613

function of the prevalence and severity of the patient safety target (eg, medical error) and the
current utilization of the practice.
Of course the actual impact of any practice is assessable only if factors related to
evidence supporting the practice are evaluated. Since the Report represents an assemblage of
the evidence for patient safety practices, the instructions to authors outlined the detailed data
elements related to study design and outcomes we required them to abstract from the relevant
studies (see Chapter 3). This information was used to assess, in general terms, the overall
strength of the studies for each practice. Effectiveness is commonly defined as the net positive
effect in routine practice. Frequently the data reported in the studies related to efficacy, usually
the net positive effect under controlled, experimental situations. The translation from efficacy to
effectiveness is not straightforward if no direct evidence is available, and is therefore based on
judgments about the generalizibility of the specific research studies conducted. Also of key
importance, and therefore abstracted from studies for use in the ratings, was the effect size of the
intervention.
Finally, evidence-based reviews consider the potential for harm from a medical
intervention, and authors were asked to report on any relevant evidence, as well as reasoned
concerns, gleaned from the literature or from common knowledge about a practice.
To address the real-world environment and the desire by the public for action in the area
of patient safety, practice chapters were designed to include information about cost and other
potential barriers to implementation. While authors sometimes discussed cost savings or
reported cost-effectiveness analyses, the focus was on the start-up costs and annual outlays for
ongoing use of the practice. Although initial and ongoing costs are a function of local
environments (eg, size of the health care network or institution), possible cost savings are likely
to be even more subject to local conditions (eg, prevalence of the patient safety target). For
major investment decisions, an assessment of tradeoffs is more appropriate at the local level.
Our intention was simply to report “ballpark” estimates of initial and recurring costs. Separate
from economic consequence of a particular practice implementation are the political and
technical considerations.
For all of these data inputs into the practice ratings, the primary goal was to find the best
available evidence from publications and other sources. Because the literature has not been
previously organized with concurrent considerations of each of these areas, most estimates could
be improved with further research and some are informed by only general and somewhat
speculative knowledge. Where possible, in the summaries of these elements, we have attempted
to highlight assessments made on the basis of limited data.
Rating Process
The 4-person Editorial Team developed a rating form that captured the patient safety
target, practice description, and general rating categories (eg, High, Medium, Low) for some of
the elements described in the section above. General heuristics were specified for each category,
although individual judgment for ratings was designed into the process. The form also specified
comment areas to allow raters to document their specific judgments and concerns about ratings.
Each chapter was independently rated by each Editor as to the practices for which there was
evidence. The Editorial Team convened for 3 days to compare scores, discuss disparities, and
come to consensus about ratings–both by category and summary ratings–or the reviewed
practices.
614

Details about Decision Rules and Judgment Considerations
Potential Impact Factor
As noted above, an assessment of potential impact considered the prevalence and severity
of the patient safety target, and the current utilization of the practice being evaluated. The
Editorial Team used the data from the chapters and clinical knowledge to order the potential
impact as “High,” “Medium,” “Low,” or “Insufficient Information.” To qualify for the “High”
score, a practice had to target a patient population of greater than 1% of hospitalized patients
(about 300,000 patients per year) or target a patient safety problem that can result in death or
disability. The “Low” score was used for target populations of less than 0.01% of hospitalized
patients (about 3000 patient/year) who might experience reversible adverse effects if an effective
practice were not available. Potential impact was deemed a “Medium” if the practice had a
patient safety target that fell between the 2 other categories.
An additional decision rule was applied to the Impact rating after the initial assessment
based on prevalence and severity was made. If a practice was currently widely used (>75% of
hospitals), then the rating was demoted one notch (ie, from High to Medium or Medium to Low).
When this situation occurred, a notation identified that the potential impact level was impacted
by its high current utilization.
We reserved the “Insufficient Information” category for those cases where the prevalence
and severity information was quite limited or where the patient safety target was ill-defined.
Evidence Supporting the Practice
Study strength, effect size on target(s), and need for vigilance due to potential harms
were rated based more on judgment than pre-specified decision rules. In each case, raters
documented their reasons for category choices.
For study strength, the level of study design and outcomes (see Chapter 3 for
hierarchies), number of studies, numbers of patients in studies, generalizability, and other
methodologic issues were specified as factors to consider in weighting the relative study strength
for a particular practice. Study strength could be categorized as “High,” “Medium,” or “Low.”
The actual findings of the studies were not considered when scoring study strength because this
information was captured in the assessment of effect size on target. If there was minimal or no
evidence about a practice, the study strength rating was “Low” and raters did not score the
remaining 2 elements of the evidence supporting the practice since that might give undue
“credit” to the findings.
The assessment of effect size on target(s) was based on the relative risk reductions or
odds ratios reported in the reviewed studies for evidence of effectiveness. The raters only used
the findings reported in the practice chapters, and did not perform additional analyses (eg, metaanalysis).
If all studies or, in cases where there were a large number of studies, the vast majority
showed a positive and appreciable effect size (ie, greater than 15% relative risk reduction), then
the positive effect size was categorized as “Robust.” If there was clearly no effect or a very
minimal effect (ie, less than 5% relative risk reduction), then the positive effect size was rated as
“Negligible.” For findings that were considered suggestive of substantive effect, but not clearly
“Robust,” the category used was “Modest.” The final category, “Unclear,” captured those
practices for which the effect size results were inconsistent.
For any given practice that reduces one adverse event, it is conceivable that new
problems might ensue when the practice is implemented. Thus, we subjectively rated the
615

concern for harm based on the level of vigilance necessary to ensure that the practice, if
implemented, would not result in collateral negative effects. The categories available were
“Low,” “Medium,” and “High.” Thus, a practice rated as “Low” would require little to no
attentiveness to potential harms, while one rated as “High” would merit heightened monitoring
for potential negative effects. These ratings were made conservatively, meaning that when in
doubt, a higher vigilance category was selected.
Implementation
Assuming a 3-year lead time for implementation, patient safety practices were rated for
their costs and complexity. Costs were based on initial start-up and annual expenditures for full
implementation at an average size hospital or health care organization. Potential cost savings
were not considered for the rating, but were reported in the practice chapters if they were
documented in the literature. If a practice was expected to require expenditures of greater than
about $1 million, the rating was “High.” Expenditures of approximately $100,000-$1 million
were categorized as “Medium.” Below this level, practices were rated as “Low” in terms of cost.
The feasibility of implementation was rated by considering potential political (eg, major
shifts in who delivers care) and technical (eg, integration of legacy and newer computer systems)
obstacles. Because relatively few data exist for rating implementation complexity, we used only
2 categories, “Low” and “High,” meaning relatively easy and relatively difficult. In cases in
which implementation could be accomplished simply with the expenditure of dollars, we gave
high cost scores but low feasibility scores.
Overall Rating for Impact/Evidence
In addition, each member of the team considered the totality of information on potential
impact and evidence supporting the practice to score each on a 0 to 10 scale (“Strength of the
Evidence”). For these ratings, we took the perspective of a leader of a large health care
enterprise (eg, a hospital or integrated delivery system) and asked the question, “If you wanted to
improve patient safety at your institution over the next 3 years and resources were not a
significant consideration, how would you grade this practice?” For this rating, we explicitly did
not consider difficulty or cost of implementation in the rating. Rather, the rating simply
reflected the strength of the evidence regarding the effectiveness of the practice and the probable
impact of its implementation on reducing adverse events related to health care exposure. If the
patient safety target was rated as “High” impact and there was compelling evidence (ie, “High”
relative study strength) that a particular practice could significantly reduce (eg, “Robust” effect
size) the negative consequences (eg, hospital-acquired infections), raters were likely to score the
practice close to 10. If the studies were less convincing, the effect size was less robust, or there
was a need for a “Medium” or “High” degree of vigilance because of potential harms, then the
rating would be lower.
Overall Rating for Research Priority
Analogously, we also rated the usefulness of conducting more research on each practice,
emphasizing whether there appeared to be questions that a research program might have a
reasonable chance of addressing successfully (“Research Priority”). Here, our “thought
question” was, “If you were the leader of a large agency or foundation committed to improving
patient safety, and were considering allocating funds to promote additional research, how would
you grade this practice?” If there was a simple gap in the evidence that could be addressed by a
research study or if the practice was multifaceted and implementation could be eased by
616

determining the specific elements that were effective, then the research priority was high. If the
area was one of high potential impact (ie, large number of patients at risk for morbid or mortal
adverse events) and a practice had been inadequately researched, then it also would also receive
a relatively high rating for research need. Practices might receive low research scores if they
held little promise (eg, relatively few patients affected by the safety problem addressed by the
practice or a significant body of knowledge already demonstrating the practice’s lack of utility).
Conversely, a practice that was clearly effective, low cost and easy to implement would not
require further research and would also receive low research scores.
Caveats to Ratings
For all elements assessed, divergent assessments among the 4 Editor-raters were
infrequent and were discussed until consensus was reached. For each final category where
differences in interpretation existed and persisted after discussion, the protocol was to document
a comment about these differences (see Chapter 59). Comments were also noted when specific
additional information could clarify concerns about fidelity of a specific rating. In a few cases,
categories that had not been specified were created for unusual circumstances and again
comments to explain the category were documented.
Rating Tables
Information Captured
Ratings were recorded on a data table, and comments were footnoted. Summarizing from
the previous discussion, the various data tables appearing in Chapters 57-59 captured some or all
of the following 11 elements and rating categories, ordered from strongest to weakest:
1. Chapter number
2. Patient safety target(s)
3. Patient safety practice description
4. Potential Impact: High, Medium, Low, Insufficient Information
5. Study Strength: High, Medium, Low
6. Effect Size: Robust, Modest, Negligible, Unclear
7. Vigilance: Low, Medium, High
8. Implementation cost: Low, Medium, High
9. Implementation complexity (political, technical): Low, High
10. Overall rating for impact/evidence: 0 to 10 (10 is highest), in 0.5 increments
11. Overall rating for research need: 0 to 10 (10 is highest), in 0.5 increments
Information Reported
Chapter 59 presents the detailed data tables with categorization of elements 1-9 above.
Reporting specific scores for the overall ratings would imply a refinement in scoring that was
neither attempted nor advisable given the nature of the information available. Where applicable,
caveats to the categorizations are appended in a series of endnotes.
In rating both the strength of the evidence and the research priority, our purpose was not
to report precise 1-10 scores, but to develop general “zones” or practice groupings. As noted
earlier, better methods are available for comparative ratings when the data inputs are available.
The relative paucity of the evidence dissuaded us from using a more precise, sophisticated, but
ultimately unfeasible, approach.
617

Chapter 57 summarizes the overall ratings for the “Strength of the Evidence" regarding
their impact and effectiveness score, and subdivides the practices into 5 zones. Practices are
listed from highest score to lowest score for each rating zone. The zones are “greatest strength”
(score of 8-10), “high strength” (score of 6-7.5), “medium strength” (4-5.5), “lower
impact/evidence scored practices” (score of 2-3.5), “lowest impact/evidence scored practices”
(score of 0-1.5). Practices near the bottom of one zone may be just as appropriate to list near the
top of the adjacent lower zone. Similarly, practices at the top of a zone may actually be more
comparable to those in the adjacent higher zone. The cut-offs between zones are somewhat
artificial, but allow a general and reasonable synthesis of the data on impact and evidence
supporting (or negating) the effects of the practice. Readers can be confident that practices that
fall in the highest zone do not belong in the lowest zone.
Chapter 58 summarizes the overall ratings for the “Research Priority” score, and provides
examples of types of research that may be helpful. Practices are categorized in 3 zones: “Further
Research Likely to be Highly Beneficial” (scores of 7 and higher), “Further Research Likely to
be Beneficial” (scores of 4 to 6.5 inclusive), and “Low Priority for Research” (below 4). The
chapter lists practices for each of the top two categories of research priority.
618

Chapter 57. Practices Rated by Strength of Evidence
After rating practices on a metric for potential impact, and on the strength of the
evidence, we grouped them into 5 categories (Tables 57.1-57.5). These categorizations reflect
the current state of the evidence. If a practice that addresses a highly prevalent or severe patient
safety target receives a low rating on the impact/evidence scale, it may be because the strength of
the evidence base is still weak due to lack of evaluations. As a result the practice is likely to
show up at a high level on the research priority scale. However, if the practice has been studied
rigorously, and there is clear evidence that its effectiveness is negligible, it is rated at the low
ends of both the “strength of the evidence” (on impact/effectiveness) scale and the “research
priority” scale.
For each practice listed in Tables 57.1 through 57.5, a designation for the cost and
complexity of implementation of the practice is included. The ratings for implementation are
“Low,” which corresponds to low cost and low complexity (eg, political, technical); “Medium,”
which signifies low to medium cost and high complexity, or medium to high cost and low
complexity; and “High,” which reflects medium to high cost and high complexity.
Several practices are not included in the tables because they were not rated. This set of
practices have long histories of use outside of medicine, but have not yet received enough
evaluations for their potential health care applications:
• Promoting a Culture of Safety (Chapter 40)
• Use of Human Factors Principles in Evaluation of Medical Devices (Subchapter 41.1)
• Refining Performance of Medical Device Alarms (eg, balancing sensitivity and
specificity of alarms, ergonomic design) (Subchapter 41.2)
• Fixed Shifts or Forward Shift Rotations (Chapter 46)
• Napping Strategies (Chapter 46)
619

Table 57.1. Patient Safety Practices with the Greatest Strength of Evidence Regarding their
Impact and Effectiveness
Chapter Patient Safety Target Patient Safety Practice Implementation
Cost/Complex
31 Venous thromboembolism (VTE) Appropriate VTE prophylaxis Low
25 Perioperative cardiac events in
patients undergoing noncardiac
surgery
16.1 Central venous catheter-related
bloodstream infections
Use of perioperative beta-blockers Low
Use of maximum sterile barriers during
catheter insertion
20.1 Surgical site infections Appropriate use of antibiotic prophylaxis Low
48 Missed, incomplete or not fully
comprehended informed consent
Asking that patients recall and restate what
they have been told during informed
consent
17.2 Ventilator-associated pneumonia Continuous aspiration of subglottic
secretions (CASS)
620
Low
Low
Medium
27 Pressure ulcers Use of pressure relieving bedding materials Medium
21 Morbidity due to central venous
catheter insertion
9 Adverse events related to chronic
anticoagulation with warfarin
33 Morbidity and mortality in postsurgical
and critically ill patients
16.2 Central venous catheter-related
bloodstream infections
Use of real-time ultrasound guidance during
central line insertion
Patient self management using home
monitoring devices
High
High
Various nutritional strategies Medium
Antibiotic-impregnated catheters Low

Table 57.2 Patient Safety Practices with High Strength of Evidence Regarding their Impact and
Effectiveness
Chapter Patient Safety Target Patient Safety Practice Implementation
Cost/Complex
18 Mortality associated with surgical
procedures
Localizing specific surgeries and procedures
to high volume centers
17.1 Ventilator-associated pneumonia Semi-recumbent positioning Low
26.5 Falls and fall injuries Use of hip protectors Low
8 Adverse drug events (ADEs) related
to targeted classes (analgesics, KCl,
antibiotics, heparin) (focus on
detection)
Use of computer monitoring for potential
ADEs
20.3 Surgical site infections Use of supplemental perioperative oxygen Low
621
High (varies)
Medium
39 Morbidity and mortality Changes in nursing staffing Medium
48 Missed or incomplete or not fully
comprehended informed consent
Use of video or audio stimuli Low
17.3 Ventilator-associated pneumonia Selective decontamination of digestive tract Low
38 Morbidity and mortality in ICU
patients
42.1 Adverse events related to
discontinuities in care
15.1 Hospital-acquired urinary tract
infection
Change in ICU structure—active
management by intensivist
Information transfer between inpatient and
outpatient pharmacy
High
Medium
Use of silver alloy-coated catheters Low
28 Hospital-related delirium Multi-component delirium prevention
program
30 Hospital-acquired complications
(functional decline, mortality)
37.4 Inadequate postoperative pain
management
Medium
Geriatric evaluation and management unit High
Non-pharmacologic interventions (eg,
relaxation, distraction)
Low

Table 57.3 Patient Safety Practices with Medium Strength of Evidence Regarding their
Impact and Effectiveness
Chapter Patient Safety Target Patient Safety Practice Implementation
Cost/Complex
6 Medication errors and adverse drug Computerized physician order entry (CPOE) High
events (ADEs) primarily related to
ordering process
and clinical decision support (CDSS)
42.4 Failures to communicate significant Protocols for notification of test results to Low
abnormal results (eg, pap smears) patients
47 Adverse events due to transportation
of critically ill patients between health
care facilities
Specialized teams for interhospital transport Medium
7 Medication errors and adverse drug
events (ADEs) related to ordering and
monitoring
Clinical pharmacist consultation services Medium
13 Serious nosocomial infections (eg, Barrier precautions (via gowns & gloves; Medium
vancomycin-resistant enterococcus, C.
difficile)
dedicated equipment; dedicated personnel)
20.4 Surgical site infections Perioperative glucose control Medium
34 Stress-related gastrointestinal bleeding H2 antagonists Low
36 Pneumococcal pneumonia Methods to increase pneumococcal
vaccination rate
Low
37.2 Inadequate pain relief Acute pain service Medium
9 Adverse events related to
Anticoagulation services and clinics for Medium
anticoagulation
coumadin
14 Hospital-acquired infections due to
antibiotic-resistant organisms
Limitations placed on antibiotic use Low
15.2 Hospital-acquired urinary tract
infection
Use of suprapubic catheters High
32 Contrast-induced renal failure Hydration protocols with acetylcysteine Low
35 Clinically significant misread Education interventions and continuous Low
radiographs and CT scans by nonradiologists
quality improvement strategies
48 Missed or incomplete or not fully Provision of written informed consent Low
comprehended informed consent information
49 Failure to honor patient preferences for Computer-generated reminders to discuss Medium (Varies)
end-of-life care
advanced directives
9 Adverse events related to
Protocols for high-risk drugs: nomograms Low
anticoagulation
for heparin
622

17.1 Ventilator-associated pneumonia Continuous oscillation Medium
20.2 Surgical site infections Maintenance of perioperative normothermia Low
26.2 Restraint-related injury; Falls Interventions to reduce the use of physical
restraints safely
Medium
26.3 Falls Use of bed alarms Medium
32 Contrast-induced renal failure Use of low osmolar contrast media Medium
623

Table 57.4 Patient Safety Practices with Lower Impact and/or Strength of Evidence
Chapter Patient Safety Target Patient Safety Practice Implementation
Cost/Complex
16.3 Central venous catheter-related
bloodstream infections
16.4 Central venous catheter-related
bloodstream infections
16.4 Central venous catheter-related
bloodstream infections
29 Hospital-acquired complications (eg,
falls, delirium, functional decline,
mortality)
37.1 Inadequate pain relief in patients with
abdominal pain in hospital patients
45 Adverse events due to provider
inexperience or unfamiliarity with
certain procedures and situations
11 Adverse drug events (ADEs) in drug
dispensing and/or administration
Cleaning site (povidone-iodine to
chlorhexidine)
624
Low
Use of heparin Low
Tunneling short-term central venous
catheters
Medium
Geriatric consultation services High
Use of analgesics in the patient with acute
abdomen without compromising diagnostic
accuracy
Low
Simulator-based training Medium
Use of automated medication dispensing
devices
12 Hospital-acquired infections Improve handwashing compliance (via
education/behavior change; sink technology
and placement; washing substance)
49 Failure to honor patient preferences for
end-of-life care
43.1 Adverse events due to patient
misidentification
10 Adverse drug events (ADEs) in
dispensing medications
Use of physician order form for lifesustaining
treatment (POLST)
Medium
Low
Low
Use of bar coding Medium (Varies)
Unit-dosing distribution system Low
24 Critical events in anesthesia Intraoperative monitoring of vital signs and
oxygenation
42.2 Adverse events during cross-coverage Standardized, structured sign-outs for
physicians
44 Adverse events related to team
performance issues
46 Adverse events related to fatigue in
health care workers
Applications of aviation-style crew resource
management (eg, Anesthesia Crisis
Management; MedTeams)
Limiting individual provider’s hours of
service
Low
Low
High
High

57.5 Patient Safety Practices with Lowest Impact and/or Strength of Evidence
Chapter Patient Safety Target Patient Safety Practice Implementation
Cost/Complex
23 Complications due to anesthesia
equipment failures
42.3 Adverse events related to information
loss at discharge
Use of pre-anesthesia checklists Low
Use of structured discharge summaries Low
22 Surgical items left inside patients Counting sharps, instruments and sponges Low
17.4 Ventilator-associated pneumonia Use of sucralfate Low
26.4 Falls and fall-related injuries Use of special flooring material in patient
care areas
43.2 Performance of invasive diagnostic or
therapeutic procedure on wrong body
part
625
Medium
“Sign your site” protocols Medium
26.1 Falls Use of identification bracelets Low
32 Contrast-induced renal failure Hydration protocols with theophylline Low
47 Adverse events due to transportation
of critically ill patients within a
hospital
16.4 Central venous catheter-related
bloodstream infections
16.4 Central venous catheter-related
bloodstream infections
Mechanical rather than manual ventilation
during transport
Low
Changing catheters routinely High
Routine antibiotic prophylaxis Medium

626

Chapter 58. Practices Rated by Research Priority
Further research on a number of practices would clarify a range of questions (eg, whether
the practice is effective, what aspects of a multi-faceted intervention matter the most, how best to
implement the practice). The conceptual framework for this categorization is described in
Chapter 56. In Table 58.1 and 58.2, the practices are grouped in zones: “research likely to be
highly beneficial,” and “research likely to be beneficial.” We also list, in the far-right column,
the practices’ categorization for “Strength of the Evidence” (as detailed above in Tables 57.1-
57.5). For presentation in this table, this category is simplified into a 1 (“highest strength of
evidence”) to 5 (“lowest strength of evidence”) which corresponds exactly to the groupings in
Tables 57.1-5. We list these here to allow the reader to compare and contrast the research
priority rankings with the evidence rankings. Practices that are not listed in either Table 58.1 or
58.2 may benefit from more research, but were not scored as highly as those included in these 2
lists.
Table 58.1 Further Research Likely to be Highly Beneficial
Chapter Patient Safety Target Patient Safety Practice Strength of
the Evidence
(1-5 Scale; 1
is highest)
20.4 Surgical site infections Perioperative glucose control 3
18 Mortality associated with surgical Localizing specific surgeries and
2
procedures
procedures to high volume centers
20.3 Surgical site infections Use of supplemental perioperative oxygen 2
39 Morbidity and mortality Changes in nursing staffing 2
15.1 Hospital-acquired urinary tract
infection
6 Medication errors and adverse
drug events (ADEs) primarily
related to ordering process
14 Hospital-acquired infections due
to antibiotic-resistant organisms
Use of silver alloy-coated catheters 2
Computerized physician order entry
(CPOE) with clinical decision support
(CDSS)
Limitations placed on antibiotic use 3
20.1 Surgical site infections Appropriate use of antibiotic prophylaxis 1
31 Venous thromboembolism (VTE) Appropriate VTE prophylaxis 1
33 Morbidity and mortality in postsurgical
and critically ill patients
37.1 Inadequate pain relief in patients
with abdominal pain in hospital
patients
Various nutritional strategies (especially
early enteral nutrition in critically ill and
post-surgical patients)
Use of analgesics in the patient with acute
abdomen without compromising
diagnostic accuracy
627
3
1
4

12 Hospital-acquired infections Improve handwashing compliance (via
education/behavior change; sink
technology and placement; washing
substance)
9 Adverse events related to chronic
anticoagulation with warfarin
21 Morbidity due to central venous
catheter insertion
38 Morbidity and mortality in ICU
patients
Patient self-management using home
monitoring devices
Use of real-time ultrasound guidance
during central line insertion
Change in ICU structure—active
management by intensivist
32 Contrast-induced renal failure Hydration protocols with acetylcysteine 3
43.1 Adverse events due to patient
misidentification
Use of bar coding 4
27 Pressure ulcers Use of pressure relieving bedding
materials
20.2 Surgical site infections Maintenance of perioperative
normothermia
25 Perioperative cardiac events in
patients undergoing noncardiac
surgery
48 Missed or incomplete or not fully
comprehended informed consent
Use of perioperative beta-blockers 1
Use of video or audio stimuli 2
28 Hospital-related delirium Multi-component delirium prevention
program
7 Medication errors and adverse
drug events (ADEs) related to
ordering and monitoring
13 Serious nosocomial infections (eg,
vancomycin-resistant
enterococcus, C. difficile)
9 Adverse events related to
anticoagulation
48 Missed, incomplete or not fully
comprehended informed consent
49 Failure to honor patient
preferences for end-of-life care
Clinical pharmacist consultation services 3
Barrier precautions (via gowns & gloves;
dedicated equipment; dedicated personnel)
Anticoagulation services and clinics for
coumadin
Provision of written informed consent
information
Computer-generated reminders to discuss
advanced directives
628
4
1
1
2
1
3
2
3
3
3
3

9 Adverse events related to
anticoagulation
Protocols for high-risk drugs: nomograms
for heparin
26.3 Falls Use of bed alarms 3
11 Adverse drug events (ADEs) in
drug dispensing and/or
administration
Use of automated medication dispensing
devices
629
3
4

Table 58.2 Further Research Likely to be Beneficial
Chapter Patient Safety Target Patient Safety Practice Impact/
Evidence
Category
630
(1-5)
17.2 Ventilator-associated Continuous aspiration of subglottic 1
pneumonia
secretions (CASS)
17.1 Ventilator-associated
pneumonia
Semi-recumbent positioning 2
26.5 Falls and fall injuries Use of hip protectors 2
30 Hospital-acquired
Geriatric evaluation and management 2
complications (functional
decline, mortality)
unit
47 Adverse events due to Specialized teams for interhospital 3
transportation of critically ill
patients between health care
facilities
transport
34 Stress-related gastrointestinal
bleeding
H2-antagonists 3
37.2 Inadequate pain relief Acute pain service 3
15.2 Hospital-acquired urinary tract
infection
Use of suprapubic catheters 3
26.2 Restraint-related injury; Falls Interventions to reduce the use of
physical restraints safely
3
45 Adverse events due to provider
inexperience or unfamiliarity
with certain procedures and
situations
Simulator-based training 4
49 Failure to honor patient Use of physician order form for life- 4
preferences for end-of-life care sustaining treatment (POLST)
42.2 Adverse events during cross- Standardized, structured sign-outs for 4
coverage
physicians
44 Adverse events related to team Applications of aviation-style crew 4
performance issues
resource management (eg, Anesthesia
Crisis Management; MedTeams)
16.2 Central venous catheter-related
bloodstream infections
17.3 Ventilator-associated
pneumonia
Antibiotic-impregnated catheters 1
Selective decontamination of digestive
tract
2

42.4 Failures to communicate
significant abnormal results
(eg, pap smears)
Protocols for notification of test results
to patients
36 Pneumococcal pneumonia Methods to increase pneumococcal
vaccination rate
3
16.3 Central venous catheter-related Cleaning site (povidone-iodine to
4
bloodstream infections chlorhexidine)
16.4 Central venous catheter-related
bloodstream infections
Use of heparin 4
16.4 Central venous catheter-related Tunneling short-term central venous 4
bloodstream infections catheters
29 Hospital-acquired
complications (eg, falls,
delirium, functional decline,
mortality)
Geriatric consultation services 4
46 Adverse events related to Limiting individual provider’s hours of 4
fatigue in health care workers service
26.4 Falls and fall-related injuryies Use of special flooring material in
patient care areas
5
43.2 Performance of invasive
diagnostic or therapeutic
procedure on wrong body part
“Sign your site” protocols 5
42.1 Adverse events related to Information transfer between inpatient 2
discontinuities in care and outpatient pharmacy
48 Missed, incomplete or not fully Asking that patients recall and restate 1
comprehended informed what they have been told during
consent
informed consent
8 Adverse drug events (ADEs) Use of computer monitoring for
2
related to targeted classes
(analgesics, KCl, antibiotics,
heparin) (focus on detection)
potential ADEs
24 Critical events in anesthesia Intraoperative monitoring of vital signs
and oxygenation
4
42.3 Adverse events related to
information loss at discharge
Use of structured discharge summaries 5
631
3

632

Chapter 59. Listing of All Practices, Categorical Ratings, and
Comments
Ch. # Patient Safety
Target
6 Medication errors
and adverse drug
events (ADEs)
primarily related
to ordering
process
7 Medication errors
and ADEs related
to ordering and
monitoring
8 ADEs related to
targeted classes
(analgesics, KCl,
antibiotics,
heparin) (focus on
detection)
9 Adverse events
related to
anticoagulation
9 Adverse events
related to
anticoagulation
9 Adverse events
related to chronic
anticoagulation
with warfarin
10 ADEs in
dispensing
medications
11 ADEs in drug
dispensing and/or
administration
12 Hospital-acquired
infections
Patient Safety
Practice
Computerized
physician order
entry (CPOE) with
clinical decision
support system
(CDSS)
Clinical pharmacist
consultation
services
Use of computer
monitoring for
potential ADEs
Protocols for high
risk drugs:
nomograms for
heparin
Anticoagulation
services and clinics
for coumadin 8
Patient selfmanagement
using
home monitoring
devices
Unit-dosing
distribution system
Use of automated
medication
dispensing devices
Improved
handwashing
compliance (via
education/behavior
change; sink
technology and
placement;
Impact Study
Strength
633
Effect Size Vigilance Cost Complexity
High Medium 1 Modest Medium High 2
High Medium Modest 3
Medium Medium Robust 4
High
Low High Low
Low Medium 5
Low
Medium Medium 6 Robust 7 Medium Low Low
High Medium Unclear Low Medium Low
High High Robust Medium Medium 9 High 10
Medium
11
High Medium
Medium Unclear Low Low Low
12
High Medium
14
Unclear Medium Medium
Unclear 15
13
Low
Low Low Low 16

Ch. # Patient Safety
Target
13 Serious
nosocomial
infections (eg,
vancomycinresistant
enterococcus, C.
difficile)
Patient Safety
Practice
washing substance)
Barrier precautions
(via gowns &
gloves; dedicated
equipment;
dedicated
personnel)
14 Hospital-acquired Limitations placed
infections due to
antibiotic-resistant
organisms
on antibiotic use
15.1 Hospital-acquired
urinary tract
infection
15.2 Hospital-acquired
urinary tract
infection
16.1 Central venous
catheter-related
blood infections
16.2 Central venous
catheter-related
blood infections
16.3 Central venous
catheter-related
blood infections
16.4 Central venous
catheter-related
blood infections
16.4 Central venous
catheter-related
blood infections
16.4 Central venous
catheter-related
blood infections
Use of silver alloycoated
catheters
Use of suprapubic
catheters
Use of maximum
sterile barriers
during catheter
insertion
Antibioticimpregnated
catheters
Cleaning site
(povidone-iodine
to chlorhexidine)
Changing catheters
routinely
Impact Study
Strength
High Medium
634
17
Effect Size Vigilance Cost Complexity
Robust Medium
High 20 Medium Modest Medium
High High Unclear 22
High High Unclear
23
18
21
Medium Low 19
Low Low
Low Low Low
Medium High High
Medium High Robust Low Low Low 24
Medium High Robust Low 25
Low Low
Medium High Unclear Low Low Low
Medium High Negligible
±
NA High High
Use of heparin Medium High Unclear Medium Low Low
Tunneling shortterm
central venous
catheters
± Actually, studies show a detrimental effect of practice.
Medium High Unclear Low Low High

Ch. # Patient Safety
Target
16.4 Central venous
catheter-related
blood infections
17.1 Ventilatorassociated
pneumonia
17.1 Ventilatorassociated
pneumonia
17.2 Ventilatorassociated
pneumonia
17.3 Ventilatorassociated
pneumonia
17.4 Ventilatorassociated
pneumonia
18 Mortality
associated with
surgical
procedures
20.1 Surgical site
infections
20.2 Surgical site
infections
20.3 Surgical site
infections
20.4 Surgical site
infections
21 Morbidity due to
central venous
catheter insertion
Patient Safety
Practice
Routine antibiotic
prophylaxis
Semi-recumbent
positioning
Continuous
oscillation
Continuous
aspiration of
subglottic
secretions (CASS)
Selective
decontamination of
digestive tract
Impact Study
Strength
635
Effect Size Vigilance Cost Complexity
Medium Medium Negligible Medium Medium Low
High Medium Robust 26
Low Low Low
High High Robust 27 Medium Medium Low
High High Robust 28
High High Robust 30 Medium
Sucralfate High High Unclear High 32
Localizing specific
surgeries and
procedures to high
volume centers
Appropriate use of
antibiotic
prophylaxis
Maintenance of
perioperative
normothermia
Use of
supplemental
perioperative
oxygen
Perioperative
glucose control
Use of real-time
ultrasound
guidance during
central line
insertion
22 Surgical items left Counting sharps,
instruments,
High Medium
Medium
34
33
High Medium
Low Low High 29
31
Low Low
Low Low
Varies Medium Varies High
High Robust Medium
36
High Medium
38
35
Robust Medium
37
Low Low
Low Low
Robust Low Low Low
High Medium Robust Medium Low High 39
High High Robust 40
Low 41 Medium High
Insuff. Low Not rated Not rated Low Low

Ch. # Patient Safety
Target
Patient Safety
Practice
inside patient sponges Info. 42
23 Complications Use of preoperative
due to anesthesia anesthesia
equipment failures checklists
24 Critical events in
anesthesia
25 Perioperative
cardiac events in
patients
undergoing
noncardiac
surgery
Intraoperative
monitoring of vital
signs and
oxygenation
Use of
perioperative betablockers
26.1 Falls Use of
identification
bracelets
26.2 Restraint-related
injuries; Falls
Interventions to
reduce the use of
physical restraints
safely
Impact Study
Strength
Low 43
Low 44 Medium
45
636
Effect Size Vigilance Cost Complexity
Low Not rated Not rated Low Low
Unclear 46
Low Low Low
High High Robust Medium Low Low
Medium Medium Negligible Low Low Low
Medium Medium Unclear 47 Medium Medium Low
26.3 Falls Use of bed alarms Medium Medium Unclear Low 48 Medium
26.4 Falls and fallrelated
injuries
26.5 Falls and fall
injuries
Use of special
flooring material in
patient care areas
Use of hip
protectors
27 Pressure ulcers Use of pressure
relieving bedding
materials
28 Hospital-related
delirium
Multi-component
delirium
prevention
program
29 Hospital-acquired Geriatric
complications (eg, consultation
falls, delirium,
functional decline,
mortality)
services
49
Low
Medium Low Not rated Not rated High Low
Medium High Robust Medium Low 50
High High Robust 52
Low 51
Low High Low
High Medium Robust Low Medium Low
High High Varies 53
Low Medium High

Ch. # Patient Safety
Target
Patient Safety
Practice
30 Hospital-acquired Geriatric
complications evaluation and
(functional
decline, mortality)
management unit
31 Venous
thromboembolism
(VTE)
32 Contrast-induced
renal failure
32 Contrast-induced
renal failure
32 Contrast-induced
renal failure
33 Morbidity and
mortality in postsurgical
and
critically ill
patients
34 Stress-related
gastrointestinal
bleeding
35 Clinically
significant
misread
radiographs and
CT scans by nonradiologists
36 Pneumococcal
pneumonia
37.1 Inadequate pain
relief in hospital
patients with
abdominal pain
37.2 Inadequate pain
relief
Appropriate VTE
prophylaxis
Use of low osmolar
contrast media
Hydration
protocols with
theophylline
Hydration
protocols with
acetylcysteine
Various nutritional
strategies
Impact Study
Strength
High High Modest 54
637
Effect Size Vigilance Cost Complexity
Low Medium High
High High Robust Medium Low Low 55
Medium High Robust Low High 56
Low
Medium High Negligible Low Low Low
Medium Medium
57
Robust Low Low Low
High High Robust 58 Medium Medium Low
H2-antagonists Medium High Unclear Medium
59
Education
interventions and
continuous quality
improvement
strategies
Methods to
increase
pneumococcal
vaccination rate
Use of analgesics
in patients with
acute abdomen
without
compromising
diagnostic accuracy
Low Low
Medium Medium Robust Low Low Low
Medium High Unclear 60 Low 61
High Medium
Acute pain service High Medium Robust 64
62
Low Low
Robust Medium Low Low 63
Low 65 Medium Low

Ch. # Patient Safety
Target
37.4 Inadequate
postoperative pain
management
38 Morbidity and
mortality in ICU
patients
39 Morbidity and
mortality
Patient Safety
Practice
Nonpharmacologic
interventions (eg,
relaxation,
distraction)
Change in ICU
structure—active
management by
intensivist
Changes in nursing
staffing
40 Any safety Promoting a
problem amenable culture of safety
to culture
41.1 Medical device
related adverse
events
Use of human
factors principles
in evaluation of
medical devices
41.2 Adverse events Refining
performance of
medical device
alarms (eg,
balancing
sensitivity and
specificity of
alarms, ergonomic
design)
42.1 Adverse events
related to
discontinuities in
care
42.2 Adverse events
during crosscoverage
Information
transfer between
inpatient and
outpatient
pharmacy
Standardized,
structured signouts
for physicians
42.3 Adverse events Use of structured
related to discharge
information loss at summaries
discharge
42.4 Failures to
communicate
significant
abnormal results
Protocols for
notification of test
results to patients
Impact Study
Strength
638
Effect Size Vigilance Cost Complexity
High High Unclear Low Low Low
High Medium Robust 66
High Medium
Insuff.
Info.
Insuff.
Info.
High 71
67
** 69
** 70
** 72
Low Medium High
Varies Low High Low 68
High Medium Robust Low Medium
Medium Low Not rated Not rated Low 74
Insuff.
Info
Varies High
Varies High
Varies High
73
Low
Low
Low 75 Not rated Not rated Low Low
Medium Medium Modest Low Low Low

Ch. # Patient Safety
Target
(eg, pap smears)
43.1 Adverse events
due to patient
misidentification
43.2 Performance of
invasive
diagnostic or
therapeutic
procedure on
wrong body part
44 Adverse events
related to team
performance
issues
45 Adverse events
due to provider
inexperience or
unfamiliarity with
certain procedures
and situations
46 Adverse events
related to fatigue
in health care
workers
46 Adverse events
related to fatigue
in health care
workers
46 Adverse events
related to fatigue
in health care
workers
47 Adverse events
due to
transportation of
critically ill
patients between
health care
facilities
Patient Safety
Practice
Use of bar coding High 76
“Sign your site”
protocols
Application of
aviation style crew
resource
management (eg,
Anesthesia Crisis
Management;
MedTeams)
Simulator-based
training
Limiting individual
provider’s hours of
service
Fixed shifts or
forward shift
rotations
Impact Study
Strength
639
Effect Size Vigilance Cost Complexity
Low Not rated Not rated Varies 77
High
High Low Not rated Not rated Low High
High 78
Insuff.
Info 79
Insuff.
Info.
Insuff.
Info.
Napping strategies Insuff.
Info.
Specialized teams
for interhospital
transport
Low Not rated Not rated Medium High
Medium
80
Unclear
81
Low Medium Low
Medium Unclear Low High High
** 82
** 84
Medium Medium
86
Varies
83
High 85
Varies
Low
Modest Low Medium Low

Ch. # Patient Safety
Target
47 Adverse events
due to
transportation of
critically ill
patients within a
hospital
48 Missed,
incomplete or not
fully
comprehended
informed consent
48 Missed,
incomplete or not
fully
comprehended
informed consent
48 Missed,
incomplete or not
fully
comprehended
informed consent
49 Failure to honor
patient
preferences for
end-of-life care
49 Failure to honor
patient
preferences for
end-of-life care
Patient Safety
Practice
Mechanical
ventilation
Asking that
patients recall and
restate what they
have been told
during informed
consent
Use of video or
audio stimuli
Provision of
written informed
consent
information
Computergenerated
reminders to
discuss advanced
directives
Use of physician
order form for lifesustaining
treatment (POLST)
Impact Study
Strength
640
Effect Size Vigilance Cost Complexity
Medium Medium Negligible Low Low Low
High Medium Robust Low Low Low 87
High Medium Modest Low Low 88
Low
High Medium Unclear Low Low Low
High Medium Robust Low Medium
89
Low
High Low Not rated Not rated Low Low 90

Comments Section
1 Medium strength of evidence for computerized physician order entry: although randomized
control trials have been conducted, findings from sophisticated “home grown” systems only 2-3
sites may not be fully generalizable. In addition, the impact of the practice on adverse events has
not been as well studied as for the non-clinical outcome, medication errors.
2 Cost of CPOE is substantially higher than for most other practices in the high cost category.
3 The impact of clinical pharmacists consultation services may be less than that of CPOE due to
logistics of screening large volumes of orders to target those most prone to error or most
consequential.
4 Estimate of effect size based on single study with limited target (only antibiotic treatments).
5 Cost influenced by whether existing computer systems are used in pharmacy services.
6
For nomogram protocols, study strength medium because the major concern (bleeding) is not
addressed in most studies.
7 Effect size greater than 15% for surrogate markers; not bleeding or clot rate.
8
Anticoagulation clinics: Both inpatient and outpatient venues studied, so some heterogeneity
among results.
9 Self-management of warfarin (coumadin): on average the cost per patient is low, but the
aggregated cost is medium from the perspective of an insurer or integrated system.
10 Higher complexity of implementation because self-management practice displaces locus of
control out of institution and may engender debate over insurance coverage. Other countries
cover this practice, but it is currently not covered by Medicare in the United States.
11
Unit-dosing is a ubiquitous practice that has surprisingly little evidence of effectiveness;
evidence is old and mixed.
12
Study strength is affected because outcomes measured are not the major outcomes of interest -
ie, ADEs.
13 The implementation ratings are related to patient safety only, but note that institutions may
also implement this practice for cost-savings due to less drug loss and better inventory control.
14
Study design for handwashing compliance practices generally had short duration of follow-up;
no randomized control trials.
641

15
Unclear effect size due to mixed results and no clear pattern in a group of heterogeneous
practices.
16 Rated as low, but this practices requires behavior change on the part of the provider.
Therefore, it may be more difficult to implement because its success largely rests on education
(see Chapter 54) and acceptance.
17
There are a number of studies of barrier precautions, but most are Level 3 study designs so the
strength is not rated as “High.”
18 Potential decrease in provider interaction with patients may cause psychological, as well as
other, effects if care from clinicians is compromised.
19 Rated as low, but this practices requires behavior change on the part of the provider.
Therefore, it may be more difficult to implement because its success largely rests on education
(see Chapter 54) and acceptance.
20
Impact upgraded from “medium” to “high” rating because of public health impact of more
antibiotic-resistant pathogens.
21 Practice requires active, ongoing monitoring and input from infection control officers to make
sure proper drugs are prescribed. Also, vigilance includes need for institution-wide monitoring of
pathogens.
22 The effect size of using silver alloy catheters is unclear: a well-done meta-analysis is positive,
showing decrease in bacteriuria, but more recent results of possibly better designed individual
studies are mixed regarding benefit. Also, the actual strength of the link, however intuitive,
between bacteriuria and clinically significant urinary tract infection is unclear.
23 Effect size of using suprapubic catheters is unclear because of some heterogeneity in studies.
Results are generally positive, but no meta-analysis yet conducted. In addition, the effect on
outcome of clinically significant urinary tract infections is also unclear.
24 Rated as low, but this practices requires behavior change on the part of the provider.
Therefore, it may be more difficult to implement because its success largely rests on education
(see Chapter 54) and acceptance.
25
With antibiotic-impregnated catheters made with minocycline, there is the theoretical risk of
increased antibiotic resistance.
26 Pneumonia outcome was significantly reduced, but mortality was not.
27 Meta-analysis of 6 randomized controlled trials showed significant and large relative risk
reduction, but 2 other randomized controlled trials showed no impact.
28
Benefit observed in prevention of ventilator-acquired pneumonia; no established benefit for
mortality.
642

29 High complexity for implementation since it requires retraining for a new practice.
30 Most benefit in reducing pneumonia and mortality occurs when both IV and topical
decontamination are used. Topical (by itself) only reduces ventilator-associated pneumonia.
However, topical carries less potential for harm (ie, antibiotic resistance).
31 Medium vigilance for harm because of public health concerns due to possible increase in
antibiotic resistance. The Center for Disease Control and Prevention (CDC) and the American
Thoracic Society (ATS) both recently reviewed this topic and did not recommend this practice.
32
If sucralfate were used because of its possible effect on reducing risk of ventilator-acquired
pneumonia, it would displace a practice that has more established benefit for GI bleeding (H2
blockers).
33 The study strength for localizing care to high volume centers is evaluated across a range of
practices. There are large variations in evidentiary base across specific practices. Evidence is not
structured to determine effect on patient safety. Although the literature includes possible
benchmarks/thresholds for volume levels for specific procedures, the evidence is related more to
quality enhancement than to improvements in patient safety.
34 Relatively high current utilization of practice reduced impact by one level.
35 Vigilance is required to monitor antibiotic overuse to prevent negative public health effects.
36 Study strength is rated as medium because randomized clinical trial data only applies to one
disease process, although may be generalizable.
37
Medium vigilance for harm: although not studied, for certain cohorts the practice may be
detrimental.
38 Study strength is rated as medium because randomized clinical trial data only applies to one
disease process, although may be generalizable.
39 Tight perioperative glucose control requires major shift in practice style, increased vigilance,
more coordination between nurses and physicians, and perhaps new policies regarding nursing
care for diabetics.
40 Effect size high, but more impressive decrease in “failed insertion attempts” than in more
clinically relevant complications. Also, there is some heterogeneity in study results, and there
are two different technologies assessed (plain ultrasound vs. US with doppler), and the results
vary.
41 Theoretical risk that additional manipulation/handling could increase infection risk; also
concern regarding impact on providers’ abilities to place catheters emergently when ultrasound
guidance is not available.
643

42 Insufficient information about retained sponges: the event is highly concerning and often
morbid when occurs, but the only data on frequency are from case reports.
43 Low potential impact because anesthesia complications are already so uncommon; also
difficult to determine impact of current use of some version of this practice (eg, low opportunity
possible due to current utilization).
44 Low potential impact because anesthesia complications are already so uncommon; also
difficult to determine impact of current use of some version of this practice (eg, low opportunity
possible due to current utilization).
45 Although there has been a very large randomized trial of pulse oximetry, other studies covered
additional aspects of intraoperative monitoring and were generally of lower study design quality.
46 Pulse oximetry study showed no benefit, but major potential methodologic problems, such as
secular trends. Complications that monitoring are designed to find are very unusual, so even a
large trial may have been under-powered to detect important effects.
47 Because the patient safety target is reduction of unnecessary restraints, there are multiple
outcomes of interest. Although reducing unnecessary restraints does not seem to increase the risk
of falls, it raises other concerns regarding disconnected IVs, elopement risk, etc., which have not
been fully evaluated.
48 Probably low, as categorized, but there is a theoretical risk that patients will not receive as
much attention from nurses and other providers.
49 Medium cost based on relatively widespread implementation of bed alarms required to impact
all patients who may potentially benefit. May also impact nursing workload and staffing needs.
50 Possibly higher cost if large numbers of patients would benefit from wearing hip protectors.
There is also the question of whether these costs are borne by system/insurers or patients
themselves.
51 Implementation complexity in the hospital may be low, but implementation outside of the
hospital might involve large educational campaign directed at patients who could benefit from
practice.
52 Studies compare a variety of special bedding materials to standard beds. Effect size for one
special bed option versus another is not known. Unclear which particular surface works best.
53 Effect size varies since heterogeneous outcomes, perhaps in part related to the variety of
interventions, some of which involved both inpatient and outpatient components.
54 Effect size varies due to heterogeneous results, which depend in part on the outcomes of
interest (ie, functional outcomes vs. mortality).
644

55 Rated as low, but this practices requires behavior change on the part of the provider.
Therefore, it may be more difficult to implement because its success largely rests on education
(see Chapter 54) and acceptance.
56 Total cost, of course, depends on the extent of utilization (eg, all patients versus only targeted
patients). Cost-effectiveness analyses demonstrate the importance of targeting appropriate
patients.
57 Outcome is level 2, only one study for N-acetylcystine.
58 Varies according to specific nutritional support practice. Robust findings for early enteral
nutrition in critically ill and post-surgical patients.
59 Vigilance for harm is medium because of potential risk of increasing ventilator-associated
pneumonia, and also because of possible overuse since high-risk groups are now better defined.
60 Depends on specific intervention; standing orders have the highest effectiveness.
61
Harm concern low, except one recent study (see Chapter 36) showed trend toward harm in
HIV-positive patients.
62 Although some studies were randomized control trials, they did not look at all clinically
relevant outcomes to ensure that practice was safe. Under-powered to assess whether diagnostic
capability not impaired.
63 Rated as low, but this practices requires behavior change on the part of the provider.
Therefore, it may be more difficult to implement because its success largely rests on education
(see Chapter 54) and acceptance.
64 Only studied for post-operative pain; may not apply more generally.
65
Some speculation that care may be fragmented when applied broadly, beyond post-operative
patients.
66 Some of the positive results may be attributable to factors other than the intervention. Concern
about underlying population changing (eg, secular trends).
67 Study strength is medium despite a number of studies, because of variation in practices (eg,
various measures of nurse staffing, models of care). Chapter was designed to generalize across
practices regarding nursing structure versus outcomes; evidence is not structured to tell effect on
patient safety, and there are no benchmarks/thresholds for nurse staffing levels.
68 Rated as low, but this practices requires behavior change on the part of the provider.
Therefore, it may be more difficult to implement because its success largely rests on education
(see Chapter 54) and acceptance.
645

69 Most evidence available outside of medicine; study strength not rated. These practices, drawn
largely from non-health care industries, were not fully rated because of their unique nature and
their relatively small evidentiary base in the health care literature.
70 Most evidence available outside of medicine; study strength not rated. These practices, drawn
largely from non-health care industries, were not fully rated because of their unique nature and
their relatively small evidentiary base in the health care literature.
71 Although alarms are ubiquitous in the hospital, it is unclear how many adverse events might
be improved by improvements in alarm systems.
72 Most evidence available outside of medicine; study strength not rated. These practices, drawn
largely from non-health care industries, were not fully rated because of their unique nature and
their relatively small evidentiary base in the health care literature.
73 Costs are shared among a variety of payors including outpatient pharmacy.
74 Cost would vary based on interventions considered—some low-tech, paper-based, or pocket
computers; higher cost for full-scale computerized systems.
75
Although one randomized trial performed, the outcomes reported were only indirectly related
to patient safety outcomes.
76 Somewhat unclear, but errors due to misidentification can be grave.
77 Cost varies based on specific system and level of implementation.
78
Impact is a function of how widely the practice can be used (ICU vs. ward teams vs. operating
room).
79
Insufficient information outside of anesthesia about volume of human factors errors amenable
to training approaches.
80
Limited studies with small numbers of participant and with different simulators lead to
concerns about generalizability.
81
Effect unclear since few studies with comparable simulators, and evaluated with mostly Level
3 outcomes.
82 Most evidence available outside of medicine; study strength not rated. These practices, drawn
largely from non-health care industries, were not fully rated because of their unique nature and
their relatively small evidentiary base in the health care literature.
83 Fixed shift may be more costly and difficult to implement than forward rotation.
646

84 Most evidence available outside of medicine; study strength not rated. These practices, drawn
largely from non-health care industries, were not fully rated because of their unique nature and
their relatively small evidentiary base in the health care literature.
85
Restructuring patient care to allow for napping while minimizing discontinuities could be
expensive.
86 Study strength is borderline-medium with three Level 3 studies.
87 Rated as low, but this practices requires behavior change on the part of the provider.
Therefore, it may be more difficult to implement because its success largely rests on education
(see Chapter 54) and acceptance.
88 Cost for video disks – assumes that off-the-shelf products exist for common procedures; would
be higher if an institution has to build its own systems.
89
Cost would be lower for health care organizations that already rely on computers for care
management.
90 Rated as low, but this practices requires behavior change on the part of the provider.
Therefore, it may be more difficult to implement because its success largely rests on education
(see Chapter 54) and acceptance.
647

648

Appendix: List of Contributors
Editorial Board
Kaveh G. Shojania, MD
Assistant Professor
Department of Medicine
University of California, San Francisco
San Francisco, CA
Bradford W. Duncan, MD
AHRQ Patient Safety Fellow
Center for Primary Care & Outcomes
Research
Stanford University
Stanford, CA
649
Kathryn M. McDonald, MM
Executive Director
Center for Primary Care & Outcomes
Research
Stanford University
Stanford, CA
Robert M. Wachter, MD
Professor and Associate Chairman
Department of Medicine
University of California, San Francisco
San Francisco, CA

Contributors
Joseph V. Agostini, MD
Robert Wood Johnson Clinical Scholar
Yale University School of Medicine
New Haven, CT
Chapters 26-30
Andrew D. Auerbach, MD, MPH
Assistant Professor of Medicine
University of California, San Francisco
San Francisco, CA
Chapters 18-20, 22-25
Dorothy I. Baker, PhD, RNCS
Research Scientist, Chronic Disease Epidemiology
Department of Epidemiology and Public Health
Associate Director, Claude Pepper Older Americans
Independence Center
Yale University School of Medicine
New Haven, CT
Chapters 26-30
David W. Bates, MD, MSc
Associate Professor of Medicine
Harvard Medical School
Chief, Division of General Internal Medicine
Brigham and Women's Hospital
Boston, MA
Chapters 6-9, 41, 42, 45, and 46
Sidney T. Bogardus, Jr., MD
Assistant Professor of Medicine
Yale University School of Medicine
Medical Director, Adler Geriatric Assessment
Center
Yale-New Haven Hospital
New Haven, CT
Chapters 26-30
650
Erica Brownfield, MD
Assistant Professor of Medicine
Emory University School of Medicine
Grady Memorial Hospital
Atlanta, GA
Chapter 37
Harold R. Collard, MD
Fellow in Pulmonary & Critical Care Medicine
University of Colorado Health Sciences Center
Denver, CO
Chapter 17
Lorenzo Di Francesco, MD
Assistant Professor of Medicine
Emory University School of Medicine
Department of Medicine
Grady Memorial Hospital
Atlanta, GA
Chapter 32
Daniel D. Dressler, MD
Instructor of Medicine
Emory University School of Medicine
Hospital Medicine Fellow
Grady Memorial Hospital
Atlanta, GA
Chapter 34
Bradford W. Duncan, MD
AHRQ Patient Safety Fellow
Center for Health Policy
Center for Primary Care & Outcomes Research
Stanford University
Stanford, CA
Chapters 45 and 46

Sharon K. Inouye, MD, MPH
Associate Professor of Medicine
Yale University School of Medicine
Investigative Medicine Program
New Haven, CT
Chapters 28 and 29
Salim D. Islam, MD
Assistant Clinical Professor
Department of Anesthesia and Perioperative Care
University of California, San Francisco
San Francisco VA Medical Center
San Francisco, CA 94121
Chapters 23 and 24
Ashish K. Jha, MD
Assistant Professor of Medicine
University of California, San Francisco
Division of General Internal Medicine
San Francisco VA Medical Center
San Francisco, CA
Chapters 45 and 46
Rainu Kaushal, MD, MPH
Instructor in Medicine
Harvard Medical School
Division of General Internal Medicine
Brigham and Women’s Hospital
Boston, MA
Chapter 6 and 7
651
Jennifer Kleinbart, MD
Assistant Professor of Medicine
Emory University School of Medicine
Atlanta, GA
Chapter 31
Sunil Kripalani, MD
Instructor of Medicine
Emory University School of Medicine
Division of General Medicine
Grady Memorial Hospital
Atlanta, Georgia
Chapter 35
Ebbing Lautenbach, MD, MPH, MSCE
Assistant Professor of Medicine and
Epidemiology
University of Pennsylvania School of Medicine
Hospital Epidemiologist, Presbyterian Medical
Center
Philadelphia, PA
Chapters 12-14
Susana B. Martins, MD, MSc
Fellow, Institute for Health Policy Studies
University of California, San Francisco
Foster City, CA
Chapter 47

Harvey J. Murff, MD
Fellow, Division of General Internal Medicine
Brigham and Women's Hospital
Boston, MA
Chapters 23, 41 and 42
Michael D. Murray, PharmD, MPH
Professor of Pharmacy
Purdue University School of Pharmacy
Scientist, Regenstrief Institute for Healthcare
Regenstrief Health Center, Sixth Floor
Indianapolis, Indiana
Chapters 10 and 11
David B. Nash, MD, MBA
Associate Dean and Director
Office of Health Policy and Clinical Outcomes
Thomas Jefferson University
Philadelphia, PA
Chapters 40, 44, 48 and 50
Steven Z. Pantilat, MD
Assistant Professor of Medicine
University of California, San Francisco
Project on Death in America Faculty Scholar
San Francisco, CA
Chapter 49
652
Laura T. Pizzi, PharmD
Project Director
Office of Health Policy and Clinical Outcomes
Thomas Jefferson University
Philadelphia, PA
Chapters 40, 44, 48 and 50
Jeffrey M. Rothschild, MD, MPH
Instructor in Medicine
Harvard Medical School
Division of General Medicine
Brigham and Women's Hospital
Boston, MA
Chapters 21 and 38
Kimberly Rask, MD, PhD
Assistant Professor of Medicine and Health Policy
Emory University Schools of Medicine and Public
Health
Grady Memorial Hospital
Atlanta, GA
Chapters 31,34, and 35
Sanjay Saint, MD, MPH
Assistant Professor of Medicine
Director, Patient Safety Enhancement Program
University of Michigan School of Medicine
Research Scientist
Ann Arbor VA Health Services Research &
Development Field Program
Ann Arbor, MI
Chapters 15-17

Jean Ann Seago, PhD, RN
Assistant Professor of Nursing
University of California, San Francisco
Research Fellow
Center for California Health Workforce Studies
San Francisco, CA
Chapter 39
Kaveh G. Shojania, MD
Assistant Professor
Department of Medicine
University of California, San Francisco
San Francisco, CA
Chapters 4, 5, 9, 10, 43, and 47
Robert Trowbridge, MD
Clinical Instructor
Department of Medicine
University of California, San Francisco
San Francisco, CA
Chapters 51-55
Robert M. Wachter, MD
Professor and Associate Chairman
Department of Medicine
University of California, San Francisco
Chief of the Medical Service
Moffitt-Long Hospitals
San Francisco, CA
Chapter 55
653
Heidi Wald, MD
Assistant Professor in Medicine
University of Pennsylvania
Director, Penn Hospital Care Group
Hospital of the University of Pennsylvania
Philadelphia, PA
Chapters 4,5 and 43
Scott Weingarten, MD, MPH
Associate Professor of Medicine
University of California, Los Angeles
Director, Health Services Research
Cedars-Sinai Health System
Beverly Hills, CA
Chapters 51-54
Mark V. Williams, MD
Associate Professor of Medicine
Director, Hospital Medicine Unit
Emory University School of Medicine
Medical Director, Center for Clinical
Effectiveness
Grady Memorial Hospital
Atlanta, GA
Chapters 31-35
Neil Winawer, MD
Assistant Professor of Medicine
Emory University School of Medicine
Grady Memorial Hospital
Atlanta, GA
Chapter 33

654

Index
abdominal pain, acute
analgesics for, 396
analgesics studies, 397, 399
academic detailing for modifying physician behavior, 595,
597
accreditation methods to promote patient safety practices,
607
accreditation organizations, 602
active error, 51
acute pain services
described, 400
studies, 401, 403
acute spinal cord injury, VTE risk, 338
ADE detection
alert monitors, 79
computerized alert monitors, 79, 80
computerized systems, studies, 83
pharmacists, physicians, and, 79
physician response to alert monitors, 81
research needs, 82
ADEs
anticoagulants and, 87
CDSS and, 590
clinical pharmacists and, 71, 74, 75
computerized detection and alert systems, 79
described, 79
incident reporting, 42
information transfer between inpatient/outpatient
pharmacies, 471
prevalence, 71
advance directives for end-of-life care, 556, 560
Agency for Health Care Policy and Research (AHCPR),
396, 406
Agency for Healthcare Research and Quality (AHRQ)
acknowledgments, 21
patient advocacy, 569
patient safety, 601
patient safety improvement, 13
American Medical Association (AMA) Code of Medical
Ethics, 546
American Society for Health System Pharmacists (ASHP),
antibiotic administration, 221
American Society of Anesthesiologists (ASA), 259, 265,
407, 604
amphotericin, topical, 193
analgesics
acute abdominal pain management, 396
acute abdominal pain management studies, 397, 399
antiemetics for PCA, 404
research needs, 398
surgeon decision making , impact on, 398
anesthesia
body temperature and, 231
intraoperative monitoring and, 265
pre-anesthesia checklists, 259
simulator-based training, 512
anesthesia crisis resource management (ACRM), 504
anesthesia equipment checklists, 466
antibiotic resistance and pneumococcal vaccines, 390
antibiotics
655
benefits for controlling nosocomial infections, 143
practices to change use, 141
prescription errors, 142
prophylaxis for SSI, 221, 224, 226
urinary catheters and, 149
anticoagulation
ADEs and, 88
inpatient services, 94
outpatient clinics, 88
patient self-management, 89, 96
research needs, 91
antiemetics for pain management
described, 404
studies, 405
ASEPSIS, 232, 235
aspirin to prevent venous thromboembolism, 333
atelectasis, 185, 234
audit and feedback for modifying physician behavior, 595,
596
Australian Incident Monitoring Study (AIMS), 43, 45, 266
bacteremia and urinary catheters, 149
bacteriuria and urinary catheters, 149
bar coding
bedside scanning, 491
described, 487
medical recordkeeping, 489
medication dispensing and administration, 488
patient identification, 487
patient identification studies, 490, 491
research needs, 492
specimen handling, 488
barrier precautions, central venous catheter insertion
contamination reduction, 165
studies, 165
barrier precautions, VRE and C. difficile
compliance vs. specific intervention, 127
described, 127
further research, 130
list of, 127
studies, 129, 132, 135, 136
Baxter ATC-212 dispensing system, 111
bed alarms for fall prevention in older patients
described, 291
studies, 292, 293
bed sores, 156
beds or mattresses for pressure ulcer prevention
described, 301
studies, 302, 304
bedsores, 301
behavior-based safety interventions, 450
Best Medicine, The, 570
beta-blockers, perioperative
described, 271
research needs, 274
studies, 272, 276
bloodstream infections, catheter-related (CR-BSI)
defined, 163
risks, 164
bronchopulmonary secretions, 185
business community and patient safety, 603

calorimetry, 359
Candida species in catheter-related infections, 163
capnography, 267
cardiac disease, heparin and warfarin monitoring, 88
cardiac events
hypothermia, 231
reducing with beta blockers, 271
care delivery and interventions, nurse staffing and, 423,
426, 430
catheter colonization, 163
catheter related infections
central venous, 167
catheter-related bloodstream infection (CR-BSI)
antibacterial coatings and antiseptic agents, 169
CHG solutions and, 175
defined, 163, 164
intravenous antimicrobial use of, 179
catheter-related infections
bloodstream, 163
central venous, 164
urinary tract, 149
Centers for Disease Control and Prevention (CDC), 42,
164, 194, 602
antibiotic administration, 221
pneumococcal vaccines, 385
central venous catheters
antibacterial coatings or antiseptic agents, 169, 174
barrier precautions during insertion, 165
CHG and PI studies, 177, 178
chlorhexidine gluconate at the insertion site, 175
heparin and, 178
infection studies, 167
placement of, 245
prolonged use, infection risks and risk reduction, 165
research needs, 183
ultrasound guidance for insertion, 246
use of, 163
children, transporting, 535, 536, 537
chlorhexidine gluconate (CHG) solution
central venous catheter insertion, 175
compared to PI, studies, 177, 178
studies, 175
chlorhexidine/silver sulfadiazine-impregnated catheters
described, 169
studies, 169, 172, 174
circadian rhythm, 519, 524
clinical decision support systems (CDSS)
CPOE and, 589
described, 589
research needs, 593
safety practice promotion and implementation, 590
studies, 590, 591
clinical pharmacists
ADEs and, 71
ambulatory settings, study results, 73
consultation services, 71
impact on ADEs, studies, 75
outpatient settings, studies, 72
roles and activities, 71
closed ICU model, 413
Clostridium difficile (C. difficile)
antibiotic studies, 142, 143
656
antibiotics and, 141
barrier precaution studies, 129
barrier precautions and, 127
relationship to VRE, 128
research needs, 131, 144
SSI and, 222
transferability, 127
Cockpit Management Attitudes Questionnaire (CMAQ),
502
computerized discharge summaries, 479
computerized physician order entry (CPOE) with CDSS
CDSS described, 589
computerized sign-out program, 476, 479
continuous intravenous insulin infusion (CII) during
surgery, 238
continuous oscillation
described, 185, 187
studies, 187, 189
cranial computed tomography (CT), 376
Crew Resource Management (CRM)
aviation application vs. health care, 504
aviation environment, 501
described, 501
effectiveness measurement tools, 502
health care environment, 504
operating room environment, 505
peer monitoring, 505
research needs, 507
studies, 503, 506
critical incident technique, 41
critical pathways
described, 581
medical information dissemination, 581
research needs, 584
studies, 582, 585
cross-coverage, 476
Crossing the Quality Chasm: A New Health System for the
21 st Century, 13
decubitus ulcers, 156, 301
deep vein thrombosis (DVT), 333, 341
delirium in older hospitalized patients
described, 307
geriatric consultation services and, 313
interventions, 307
research needs, 309
studies, 308, 310
diabetes
glucose control during surgery, 237
RCIN and, 349
diarrhea, C. difficile-associated, 128
direct mail patient notification, 483
discharge summaries
computerized, 479
research needs, 481
structured vs. unstructured format, 479
studies, 480, 481
disseminating medical information, 581
distal DVT, 333
domains, defined, 29
duplex ultrasonography, 334
durable power of attorney for health care (DPOA-HC), 556
educational techniques to modify physician behavior

described, 595
research needs, 598
studies, 596, 599
elastic stockings (ES) to prevent venous thromboembolism,
333
elderly
beta-blockers and, 273
delirium, preventing, 307
falling, preventing, 281
falls, preventing, 283
geriatric consultation services, 313
pneumococcal infections, 386
pneumococcal vaccines and, 387
pressure ulcers, preventing, 301
preventable deaths and functional decline, 324
embolism to the pulmonary vasculature, 333
end-of-life care
administrative initiatives, 560
advance directives, 560
described, 556
documenting patient preferences, 557
existing programs/contributions, 562
forms to promote implementation of patient preferences,
559
implementation of patient preferences, 557
improving physician/patient communication, 560
physician/patient communication, 558
preferences in outpatient setttings, 560
public education, 562
research needs, 563
enteral nutrition (EN)
immune-enhanced, 363
paths for administering, 359
studies, 361
environmental modifications to prevent delirium, 307
equipment, anesthesia, checklists for, 259
error mode prediction, 26
examination equipment, dedicated or disposable, as a
barrier precaution, 127, 136
extended-spectrum beta-lactamase (ESBL), 128, 143
falls in hospitalized and institutionalized older patients
bed alarms and, 291
geriatric consultation services and, 313
hip protectors and, 295
identification bracelets and, 284
interventions, list of, 283
physical restraints and, 287
problem description, 281
research needs, 283, 284, 289, 292, 294, 297
risk assessment, 282
risk factors, 281
special flooring and, 293
fatigue
lighting at work, 524
medical errors and, 519
medical therapies, 525
napping, 524
research needs, 526
shift limits and error reduction, 521
shift rotation and, 520
shift rotation, direction and speed, 523
sleep hygiene education, 523
657
studies, 522
fibrinogen leg scanning, 334
Flight Management Attitudes Questionnaire (FMAQ), 502
flooring, special, for fall prevention in older patients
described, 293
studies, 294, 295
Foley catheters, 149
Food and Drug Administration (FDA), 42, 51, 259, 262,
459, 489
gastroesophogeal reflux, 185
gastrointestinal bleeding
described, 368
nosocomial pneumonia and, studies, 372
research needs, 371
studies, 369
sucralfate and, 199
gentamicin, topical, 193
geriatric consultation services
described, 313
research needs, 316
studies, 314, 315, 318
geriatric evaluation and management (GEM) units
described, 313, 323
research needs, 326
studies, 324, 327
gloves/gowns as barrier precaution, 127
glucose control, perioperative
described, 237
research needs, 240
studies, 238, 241
surgical infections and, 238
government agencies and patient safety practices, 602
gram negative rods in catheter-related infections, 163
H2-receptor antagonists, 198, 368, 369, 372
hand disinfection, 119
handwashing
compliance difficulties, 119
relationship to infection, 119
research needs, 122
risk factors, 120
skin irritation to health care workers, 121
strategies to improve compliance, 121
studies, 120, 123
Health Care Financing Administration (HCFA)
medical errors, 601
physical restraints and, 288
pneumococcal vaccines, 287, 385
health literacy problem, 547
heart failure
ADEs and, 72
clinical pharmacists and, 73
heparin
about, 87
bleeding during therapy with, 88
central venous catheters and, 178
dosing and monitor errors, reducing, 87
LDUH/LMWH to prevent venous thromboembolism,
333
monitoring, 88
heparin nomograms
described, 87
PTTs and, 89

studies, 88, 92
high reliability theory, 26, 447
high-volume centers, localizing care to
Leapfrog Group’s efforts, 208
relationship to patient outcomes, 205
research needs, 208
studies, 206, 210
hip fractures
falls in older patients, 281
VTE risk in surgery, 336
hip protectors for fall prevention in older patients, 296, 298
hip replacement surgery, VTE risk, 336
histamine-2 receptor blocking agents (H2-receptor
antagonists)
described, 368
studies, 369, 372
VAP prevention, 198
hospitalists and end-of-life care, 561
human error identification, 26
human factors engineering (HFE)
anaesthesia equipment checklists, 466
described, 459
medical device alarms, 462
medical devices and errors, 459
research needs, 467
hydration protocols to prevent RCIN, 349
hygenic hand rub, 119
hypercholesterolemia, clinical pharmacists and, 73
hypertension, clinical pharmacists and, 73
hypoxemia and anesthesia, 267
identification bracelets for fall prevention in older patients
study, 285
using, 284
incident reporting
AIMS data vs. JCAHO data, 45
barriers to linking data with outcome, 46
critical incident monitoring, 41
event appropriateness, 43
intensive care units (ICUs), 42
JCAHO surveillance mandates, 43
medication errors and, 41
physician reporting, 45
research needs, 46
sentinel event alerts, 47
industries outside health care
aviation, 259, 501
behavior-based safety programs, 450
corporate cultures, 448
critical pathways, 581
disasters contributed to by fatigue, 520
incident reporting systems, 41, 42
safety cultures, 449
simulator training, 511
infection control officers, 602
information transfer, improving
abnormal results notification, 482
discharge summaries, 479
inpatient/outpatient pharmacies, 471
research needs, 474, 486
sign-out systems, 476
informed consent
658
asking patient to recall discussion information, studies,
550
form readability studies, 549
forms inadequacies, 548
legal and ethical obligations, 546
physician/patient discussion, studies, 549
procedural inefficiences in obtaining, 547
process, 546
research needs, 552
visual or auditory aids, studies, 550
written information to patients, studies, 551
Institute of Medicine (IOM) reports, 13, 205, 207, 210
intensive care units (ICUs)
incident reporting and, 42
intensivist model studies, 418
Leapgrog Group recommendations, 413
nutritional support for patients, 359
research needs, 416
structure model studies, 414, 415
structure models, 413
intensivist co-management ICU model, 413
intensivists in ICUs, 413
interhospital transportation of critically ill patients, 536,
537
interhospital transportation of critically ill patients, 534
intermittent pneumatic compression (IPC) to prevent
venous thromboembolism, 333
International Classification of Diseases (ICD-9) codes, 79
intrahospital transportation of critically ill patients, 534,
538
ischemia, myocardial, 267, 271
Joint Commission on the Accreditation of Healthcare
Organizations (JCAHO)
pain management guidelines, 396
patient safety advocacy, 602
physical restraint mandates, 288
root cause analysis mandate, 51
safety standards, 606
sentinel event alerts, 47
Sentinel Event Database, 43
surveillance mandate, 43
wrong-site surgery, avoiding, recommendations, 495
knee replacement surgery, VTE risk, 336
laparoscopic cholecystectomy training
complications with procedure, 214
learning curve, 213
research needs, 216
skills needed, 213
studies, 215
latent error, 51
Leapfrog Group
high-volume centers, 205
ICUs model recommendations, 413
methods to promote patient safety practices, 607
patient safety advocacy, 603
referral system, 208
learning curves for new procedures, 213
legal issues
end-of-life care, 556
informed consent, 546
legislation
methods to promote patient safety practices, 607

Natural Death Acts, 556
nursing issues, 425
Omnibus Budget Reconciliation Act of 1987, 288, 302
patient safety, 601
Patient Self-Determination Act (PSDA), 556
publication of performance data, 605
Lerner’s Consumer Guide, 569
lighting and fatigue, 524
LINE/LOS checklist, 502
living will for end-of-life care, 556
localizing care to high-volume centers, 205
low dose unfractionated heparin (LDUH) to prevent VTE,
333, 335
low molecular weight heparin (LMWH) to prevent VTE,
333, 335
machine-readable automatic identification (ID) systems
conventional wrist band use, 489
described, 487
medical record keeping, 489
medication dispensing and administration, 488
patient identification, 487
patient identification studies, 490, 491
research needs, 492
specimen handling, 488
malnutrition in hospitalized patients, 359, 363
McLaughlin dispensing system, 111
median sternotomy, 238
medical device alarms, 462
frequency reduction, 465
research needs, 466
source identification, 463
urgency ranking, 463
visual interfaces, 463
medical devices
design and development, 459
device evaluation, 461, 462
medical recordkeeping and bar codes, 489
medical records, patient access to, 570
medication dispensing and bar codes, 488
medication dispensing systems, automated
Baxter ATC-212 system, 111
described, 111
harmpotential, 113
machine-readable code, 111
McLaughlin system, 111
research needs, 114
studies, 111, 112, 115
unit-dose distribution systems and, 112
medication errors
clinical pharmacists impact on, studies, 75
dosage errors, 112
identification of ADEs, 41
information transfer between inpatient/outpatient
pharmacies, 471
studies, 106
time-constraints and, 101
unit-dose dispensing, 102
MedTeams behavior-based teamwork system, 505
melatonin supplementation for fatigue, 525
methicillin-resistant Staphylococcus aureus (MRSA), 128
minocycline/rifampin-coated catheters, 169
minocycline-coated urinary catheters, 149
659
misidentification, avoiding
bar coding, 487
wrong-site surgery, 494, 496, 497
wrong-site surgery, research needs, 498
monitoring, intraoperative
ASA standards, 268
described, 265
research needs, 267
studies, 266
napping, 524
National Quality Forum (NQF), 21
National Vaccine Advisory Committee, pneumococcal
vaccines, 385
Natural Death Acts, 556
nausea due to PCA, 404
neurosurgery, VTE risk, 337
non-radiologists
radiograph interpretations, 376
training, 377, 381
normal accident theory, 447
normothermia, perioperative
described, 231
studies, 232, 233
nosocomial infections
antibiotics and, 141
barrier precautions and, 127
handwashing and, 119
handwashing practices and, 120
pathogens other than VRE and C. difficile, 128
pneumonia and GI bleeding, 368
nurse staffing
availability studies, 426, 431
care delivery models, 430
care delivery studies, 426
care practices evaluated, 424
intervention studies, 427
measures, 430
nurse intervention and patient outcome studies, 437
organization models and patient outcome studies, 435
relationship of models to patient outcomes, 423
research needs, 428
studies, 425
unions and, 425
nutritional support practices
described, 359
studies, 360, 365
Occupational Safety and Health Administration (OSHA),
601
open ICU model, 413
opinion leaders for modifying physician behavior, 595, 597
opioids, 404
orientation techniques to prevent delirium, 307
oropharyngeal secretions, 190
osomolar iodinated contrast media to prevent RCIN, 349
oxygen, perioperative
described, 234
future research needs, 236
studies, 235, 236
pain management
abdomen, acute, 396
about, 396
acute pain services, described, 400

acute pain services, studies, 401, 403
antiemtics for PCA, 404
non-pharmacologic intervention studies, 407, 409
non-pharmacologic interventions, 406
patient education, 407
research needs, 408
palliative care programs, 561
patient advocacy
benefits, 569
improving non-compliance with medical advice and
treatment, 570
medical records, access to, 570
research needs, 571
patient classification systems (PCSs), 423
patient mobility to prevent delirium, 307
patient notification of abnormal results
direct mail, 483
problems and current methods, 482
research needs, 484
study, 483, 485
Patient Self-Determination Act (PSDA), 556
patient to nurse ratios, 424
peer monitoring, 505
percutaneous transluminal angioplasty, 271
pericatheter skin inflammation, 163
pharmacies
automated drug dispensing, 102
information transfer, 471, 473, 475
pharmacologic measures to prevent venous
thromboembolism, 333
physical restraints for fall prevention in older patients
described, 287
studies, 288, 290
physician behavior, modifying
education techniques, research needs, 598
education techniques, studies, 596, 599
educational techniques, 595
practice guidelines, 576
pneumococcal infections
re-infection risks, 385
vaccines and, 385
pneumococcal vaccines
antibiotic resistance, 390
delivery method studies, 392
delivery methods, studies, 388
described, 385
effectiveness studies, 386, 387, 391
research needs, 390
pneumothorax, 247
POLST (Physician Orders for Life-Sustaining Treatment),
559, 564
polymixin, topical, 193
povidone-iodine (PI)
central venous catheter insertion, 175
compared to (CHG), studies, 177, 178
practice guidelines for modifying physician behavior
computer-based, 577
described, 575
research needs, 578
studies, 576, 577
pre-anesthesia equipment checklists
described, 259
660
research needs, 261
studies, 260, 262
pressure ulcers
in older patients, 301
research needs, 304
suprapubic catheters and, 156
professional societies and patient safety, 604
professional societies, to promote patient safety practices,
607
psychosocial support for pain management, 407
public reporting systems, 605
pulmonary embolization (PE), 333
pulse oximetry, 265, 266, 466, 604
Pyxis Medstations, 112
radiocontrast-induced nephropathy (RCIN)
described, 349
prevention practices, 349
prevention studies, 353, 354
studies, 350, 351
radiograph interpretation errors by non-radiologists
described, 376
digitized vs. film-based radiographs, 376
reducing, 377, 378, 381
research needs, 380
radiologists and non-radiologists
radiograph interpretations, 376
training, 377
reminder systems for modifying physician behavior, 595
renal failure, RCIN and, 349
report methodology
abstraction elements, 34
evidence review, 15
exclusion criteria, 31
exclusions, 16
inclusion criteria, 31
inclusions, 16
organization, 18
outcome measures, 32
review methodologies, 29
safety practice ratings, 613
safety practices evaluation, 33
safety practices organization, 29
search strategy, 30
study design hierarchy, 31
review methodologies, error-based, 29
rifampin-coated urinary catheters, 149
risk assessment to prevent falls in older patients, 282
root cause analysis (RCA)
industrial accidents, 51
JCAHO mandate, 51
process, 52
research needs, 55
studies, 53, 54
safety culture, promoting
accident theories, 447
behavior-based programs, 450
checklist of patient safety concepts, 453
methods, medical and non-medical, 448
research needs, 451
VHA initiative, 449
safety practice ratings
data inputs, 613

decision rules and judgment considerations, 615
goals, 613
impact and evidence overall rating, 616
rating process, 614
rating tables, 617
research priority overall rating, 616
safety practices
all practices, categorical ratings, and comments, 633
clinical vs. general issues in content, 17
definition and scope, 13, 29
further research likely to be beneficial, 630
further research likely to be highly beneficial, 627
greatest strength of evidence regarding their impact and
effectiveness, 620
high strength of evidence regarding their impact and
effectiveness, 621
hospital care and, 14
industries outside health care, 25
low strength of evidence regarding their impact and
effectiveness, 624
lowest strength of evidence regarding their impact and
effectiveness, 625
medium strength of evidence regarding their impact and
effectiveness, 622
organization of, 29
rated by strength of evidence, 619
research priority ratings, 627
selective digestive tract decontamination (SDD)
described, 193
studies, 194, 196
semi-recumbent positioning
described, 185
studies, 186, 189
Sentinel Event Database, 43
shift rotation and fatigue, 523
sign-out systems for cross-coverage
described, 476
research needs, 478
studies, 477, 479
silicone urethral catheters, 149
silver alloy catheters
bacteriuria reduction, 151
described, 149
simulator-based training
advantages, 511
anesthesia, 512
cardiology, 514
CRM and, 512
gastroenterology, 514
radiology, 513
research needs, 514
risks, 515
surgery, 513
sleep deprivation
lighting at work, 524
medical errors and, 519
medical therapies speed, 525
napping, 524
research needs, 526
shift limits and error reduction, 521
shift rotation and, 520
shift rotation direction and speed, 523
661
sleep hygiene education, 523
studies, 522
sponge, sharp, and instrument counts
described, 255
research needs, 256
studies, 256
staff cohorting as a barrier precaution, 127
staphylococci and catheter-related infections, 163
sternal wound infections, 238
streptococcus pneumoniae, 385
stress ulcers
H2-receptor antagonist therapy, 198
studies, 201, 202
subclavian-vein catheters, 246
subglottic secretions, continuous aspiration of
described, 190
diagram, 192
studies, 193
sucralfate
gastrointestinal bleeding reduction, described, 368
gastrointestinal bleeding reduction, studies, 369, 372
studies, 201, 202
VAP reduction, 198
suprapubic catheters
compared to standard catheters, 156
compared to urethral, studies, 159
placement of, 155
research needs, 157
studies, 157
surgical site infections (SSI)
antimicrobial prophylaxis and, 221
blood flow and body temperature, 231
glucose control for diabetics and, 237
normothermia and, 232
oxygen and, 234
research needs, 223, 233
studies, 222, 224, 226
surgical sponges and instruments, retained, 255
teicoplanin and central venous catheters, 179
teleradiology, 376
thromboembolism
heparin and warfarin monitoring, 88
preventions, 333
To Err is Human: Building a Safer Health System, 13, 25
tobramycin, topical, 193
total parental nutrition (TPN), 359, 361
training
end-of-life care, for physicians, 560
laparoscoopic cholecystectomy, 213
nurses, 424
radiograph interpretations, 377
simulator-based, 511
transesophageal echocardiography, 265
transfusion errors, 489
transport teams, 534
transportation of critically ill patients
about, 534
interhospital, 534
intrahospital transport, 534
manual vs. mechanical ventilation, studies, 541
transport teams vs. standard, studies, 540
transporting children, 535, 537

transurethral resection of the prostate, 157
trauma patients, VTE risk, 338
UCSF-Stanford Evidence-based Practice Center (EPC), 13
ultrasound guidance of central vein catheterization
described, 245
insertion success rates, 246
research needs, 249
studies, 246, 247, 250
unions and nurse staffing, 425
unit-dose dispensing
automated dispensing systems, 112
described, 101
dispensing variations, 101
JCAHO standards, 102
research needs, 105
safety targets, 102
studies, 103, 104, 106
urethral catheters
compared to suprapubic, 159
latex, 149
patient discomfort, 156
research needs, 157
silicone, 149
urethral strictures, 157
urinary catheters
impregnated with antibiotic, 149
silver alloy catheters, 149
studies, 150, 153
urinary tract infections (UTIs)
catheters and, 149
prevalence, 149
research needs, 151
usability testing, 461
vancomycin, systemic use with central venous catheters,
179
vancomycin-resistant enterococci (VRE)
antibiotic studies, 142, 143
antibiotics and, 141
barrier precaution studies, 129
barrier precautions and, 127
C. difficile relationship, 128
research needs, 131, 144
transferability, 127
venography, 334
662
venous thromboembolism (VTE)
described, 333
preventative measures, mechanical and pharmacologic,
343
preventions, 333
prophylaxis for surgical procedures and medical
conditions, 345
risk in general surgery, 335
risk in hospitalized medical patients, 339
risk in neurosurgery, 337
risk in orthopedic surgery, 336
risk reduction methods, 344
studies, 334
therapy factors, 337
urethral catheters and, 156
ventilation when transporting critically ill patients, 538,
541
ventilator-associated pneumonia (VAP)
continuous aspiration of subglottic secretions, 190
continuous oscillation, 185, 187
described, 185
patient positioning studies, 189
research needs, 195, 199, 203
SDD, 193
semi-recumbent positioning, 185, 186
sucralfate, 198
Veterans Health Administration (VHA), 449
vomiting due to PCA, 404
ward stock system, 102
warfarin
about, 87
dosing and monitoring errors, reducing, 87
monitoring, 88
self-management, studies, 89, 96
venous thromboembolism prevention, 333
warfarin nomograms, studies, 92
warming blankets for normothermia during surgery, 231
web-based health information, 16, 292, 451, 561, 570
web-based programs, 478
wrong-site surgery, avoiding
described, 494
research needs, 498
studies, 496, 497